ISSN 0252-1075 Contribution from IITM

Research Report No. RR-122

Climatic and Hydroclimatic Features of Wet and Dry Spells and their Extremes across India

Nityanand Singh and

Ashwini A. Ranade

AUGUST 2009

Indian Institute of Tropical Meteorology Dr. Homi Bhabha Road, Pashan Pune - 411 008

Maharashtra, India

E-mail : lip@tropmet.res.in Fax : 91-020-25893825 Web : http://www.tropmet.res.in Telephone : 91-020-25904200

 

CONTENTS

Abstract

1. Introduction 1 2. Data used and division of the country into 19 sub-regions 3 3. Objective criterion for identification of wet and dry spells 4 4. Results and discussion 6 a. Climatological features of the wet spell 6 b. Climatological features of the extremes of wet spell 7 1) Extreme wet spell in respect of rainfall amount

2) Extreme wet spell in respect of rainfall intensity 3) Extreme wet spell in respect of duration

c. Climatological features of the dry spell 8 d. Climatological features of the extremes of dry spell 9 1) Extreme dry spell in respect of rainfall amount 2) Extreme dry spell in respect of rainfall intensity 3) Extreme dry spell in respect of duration e. Fluctuation features of the wet and dry spells and their extremes 11 f. Limitations 12 5. Summary and conclusion 13 Acknowledgements 14 References 15 Tables 17 Figures 26

 

Climatic and Hydroclimatic Features of Wet and Dry Spells and their Extremes across India

Nityanand Singh and Ashwini A. Ranade

ABSTRACT: Characteristics of wet spells (WSs) and intervening dry spells (DSs) are the most

useful information in water-related sectors. The information assumes greater significance in the wake of global climate change background and climate change scenario projections. Features of 40 parameters of two attributes wet and dry spells of the rainfall time-distribution as well as their extremes have been studied for 19 sub-regions across India using one-degree lat/lon grid cells daily rainfall data available for the period 1951-2007. In the low-frequency mode intra-annual rainfall variation WS (DS) is identified as ‘continuous period with daily rainfall equal to or greater than (less than) daily mean rainfall (DMR) of climatological monsoon period over the area of interest’. The DMR shows large spatial variation from 2.6mm/day over extreme southeast peninsula (ESEP) to 20.2mm/day over southern central West Coast (SCWC). Climatologically, number of wet spells (WSs) across the country varies from 4 over northwest India (NWI) to 11 over the ESEP and total duration 29-101 days following similar spatial pattern. Total rainfall of the WSs is however lowest (283.7mm) over the NWI (driest area) and highest (2130.4mm) over the SCWC (wettest area), similar to annual/monsoonal rainfall pattern. For individual WS duration varies from 7 to 12 days, rainfall amount 72.8-426.1mm and rainfall intensity 7.6-36.9mm/day. The rainfall due to WSs contributes 47-79% to the respective annual total. The first WS starts the earliest on 20 March over the ESEP and the latest on 21-23 June over the NWI. The last WS ends on 12 September from the NWI and on 16 December from the ESEP following normal withdrawal pattern of the summer monsoon. Inter-annual variability of starting date of the first WS along the West Coast and ending date of the last WS along the East Coast is low compared to other parts of the country due to orographic effects.

Total duration of all intervening DSs (3-10 in number) varies across the country from 45 days (WIGP) to 173 days (ESEP) and total rainfall from 27.5mm (NWI) to 553.1mm (NEI). Duration of individual DS varies from 10.5 days (central Indo-Gangetic Plain; NCIGP) to 19.9 days (NWI), and rainfall amount from 8.9mm (NWI) to 114.9mm (SCWC) and rainfall intensity 0.6-9.8mm/day following the same pattern. The contribution of rainfall due to DSs to the respective annual total varies from 9.5-26.8%. For the country as a whole, the total rainfall of WSs and DSs contributes ~85% to the annual total. Important features (total rainfall, rainfall intensity, duration and starting date) of extremes of wet as well as dry spells in respect of rainfall amount, rainfall intensity and duration have also been examined. In recent years/decades, actual and extreme WSs are slightly shorter and their rainfall intensity higher over majority of the sub-regions while actual and extreme DSs slightly longer and their rainfall intensity weaker. However, except in few isolated cases these tendencies are not found statistically significant. There is a tendency for the first WS to start ~6 days earlier across the country and the last WS to end ~2 days earlier but this shift in dates of the rain spells is also mostly not significant.

1

 

1. Introduction Rainfall is a seasonal phenomenon in tropical monsoonal climate and it occurs in spells. Start and end of rainy season; frequency, rainfall amount, rainfall intensity and duration of wet spells; and duration and severity of intervening (between two rain spells) dry spells are characterized by large spatial and temporal variations. Climatology and variability of the parameters of rainy season and wet and dry spells are valuable information for scientists, engineers, planners and managers working in water-related sectors (agriculture, ecology, hydrology and water resources). Determination of yearwise starting and ending dates of the rainy season and identification of wet and dry spells in the rainfall time distribution is the crucial issue of the problem. In this study we attempt to demonstrate that demarcation of start and end of rainy season, and identification of wet/dry spells are fundamentally one problem and single objective criterion can serve both the purposes. In the country rainy season, period of monsoon and wet season are synonymously used. During World War II, normal onset and withdrawal dates at 180 stations over Indian subcontinent (present day India, Pakistan, Bangladesh, Myanmar and Sri Lanka) were determined by the India Meteorological Department (IMD) by applying a subjective approach ‘characteristic monsoon rise/fall in 5-day period (pentad)’. For a particular station normal pentad rainfall for full calendar year (73 pentads) was obtained. The middle date of the pentad that showed characteristic monsoon rise (fall) in rainfall from previous pentad was taken as normal onset (withdrawal) date. Similarly the dates were obtained for other stations. The dates were charted and isolines of normal onset (withdrawal) dates drawn (IMD, 1943). In this historically important program, the meaning of the ‘characteristic monsoon rise/fall in pentad’ and description of the ‘other factors were also considered where the pre-monsoon thunderstorm rains merged with monsoon rain, and the transition was gradual’ are not available in the texts. But this is the most often referred work of the IMD. Raman (1974) used daily rainfall and determined year-wise (1901-1970) commencement of sowing rains at 231 locations across Maharashtra State (northwest peninsula) by applying an objective criterion: ‘A spell of at least 25mm of rain in a period of 7 days with 1 mm or more on any five of these seven days’. Normally, the first wet spell along West Coast (Konkan area) starts between 2 and 14 June, on the leeside of the West Coast (Madhya Maharashtra and Marathwada area in central peninsula) between 8 June and July, and over eastern Maharashtra (Vidarbha area in central India) between 18 June and 26 June. Standard deviation of start of first wet spell is 8 days along the West Coast, 8 to 16 days over Vidarbha area and 10 to 30 days over leeside area. He further identified second wet spell by applying the same objective criterion and calculated duration of dry spell between the two wet spells. The second wet spell occurred 8-10 days after the first wet spell over Konkan, after 10-16 days over Vidarbha and after 12-32 days over Madhya Maharashtra and Marathwada. Ananthakrishnan and Soman (1988) suggested an objective criterion to determine onset date of the summer monsoon over Kerala State (extreme southwest peninsula), the Gateway for monsoon to India; ‘when rainfall on the day and mean rainfall in the following 5 days period exceeded 10 mm’. Earlier Ananthakrishnan et al. (1967) suggested a criterion for declaring onset over Kerala, ‘beginning from 10 May if at least five out of the seven stations report 24-hourly rainfall 1 mm or more for two consecutive days the forecaster should declare on the second day that the monsoon has advanced over Kerala’. Recently, Dash et al (2009) have applied the IMD criteria of rainy day (rainfall equal to or greater than 2.5mm per day) to identify wet and dry events across India

2

 

Parts of India experience considerably longer duration of rainy season than the summer monsoon period as they get considerable rainfall from other systems also in the pre- and post-monsoon periods: Western Himalaya (Uttarakhand, Himachal Pradesh and Jammu & Kashmir States) get rainfall from western disturbances both in the pre- and post-monsoon periods, northeast India from thunderstorms both in the pre- and post-monsoon periods, and south peninsula (south of 18°N) from thunderstorms in the pre-monsoon period and from northeast monsoon in the post-monsoon period. Singh (1986) has identified climatological period of reliable rainfall (rainy season), irrespective of the system(s) that produced the rainfalls, at 466 stations across the country by applying a unified criterion, ‘continuous period with each of the monthly rainfall greater than 50mm’. Recently, Ranade et al. (2008) applied this criterion to determine yearwise start and end of the hydrological wet season (HWS) over 11 major and 36 minor river basins as well as the West Coast Drainage System (WCDS) and studied variability of the HWS parameters (starting and ending dates and duration, seasonal rainfall/rainwater and surplus rainfall/rainwater potential) over the period of longest available instrumental records (1813-2006).

In some other countries also the researchers have attempted different criteria to study rainy season and wet/dry spells. To identify wet spells across China, Bai et al. (2007) used the criteria: i) In a wet spell, the number of rainy days (rainfall ≥ 0.10 mm/day) should not be less than 4 days, and the first three days should not include a dry day. ii) A wet spell is over whenever there are two consecutive dry days. The duration of a wet spell is defined as the number of days between the two consecutive dry days. They have also documented different definitions attempted earlier in different geographical locations of China. A sophisticated hydrological scheme, the available water resources index (AWRI) has been developed by Byun and Wilhite (1999) for detecting the onset, end, and accumulated stress of drought. In the AWRI, the accumulated precipitation, the duration of accumulation and the daily reduction of water by runoff, evapotranspiration, infiltration etc. are taken into account quantitatively. Using this scheme, the onset and ending dates, and intensities of the three rainy seasons in Korea were determined for each year and each station, and also climatologically (Byun and Lee, 2002). For Peninsular Malaysia, Deni et al. (2008) defined a wet day with rainfall of at least 0.1 mm. A wet (dry) spell is defined as a period of consecutive days of exactly, say x, wet (dry) days immediately preceded and followed by a dry (wet) day. Aviad et al. (2004) studied the variation in beginning, end, and length of the rainy season along a Mediterranean-arid climate transect with the change in rainfall threshold. They changed the thresholds from 0.1 to 80 mm with interval of 2.5 mm, and from 80 to 200 mm with intervals of 5 mm, and found that as the threshold increased the rainy season began later, ended earlier, and consequently its length was shortened. Cook and Heerdegen (2001) defined rainy season as that period when the probability of 10-day dry spells was less than 0.5, and the wet season (monsoonal influence) as that period within the rainy season when the probability of dry spells was less than 0.1. A dry-day was defined with rainfall less than 5 mm. This definition was essentially meant for ecological purposes. Stern (1981) defined the start of the rainy season as the first occurrence of 20 mm of rain within 2 successive days.

In recent years/decades, declining tendency in annual rainfall is seen over 68% area of the country. Percentage area under recent negative tendency in seasonal rainfall is: winter 47%, summer 17%, summer monsoon 75% and post-monsoon 38% area; and in summer monsoon monthly rainfall: June 41%, July 85%, August 58% and September 69% area (Sontakke et al., 2008a). Dai et al. (1998) have found increases in the joint areas of the world affected by either drought and/or excessive wetness, and the results are consistent with those expected from increased green-house gases in the atmosphere.

3

 

Understanding climatic and hydroclimatic features of wet and dry spell is essential for effective agricultural and hydrological operations. On the face of global climate change background and scenario projections the problem assumed greater importance, therefore attempts have been made in recent decades world-wide to understand the problem on regional/local scales. This comprehensive study of the wet and dry spells (including their extremes) across India is undertaken with the following main objectives.

1) to develop dynamic objective criteria for identification of year-wise wet and dry spells

in different rainfall regimes of India; 2) to study detailed climatological and fluctuation features of the wet and dry spells

across the country; and 3) to examine important features of extremes of wet and dry spells.

Start of the first wet spell is expected to provide reliable estimate of starting date of

seasonal (hydrological) rainfall and end of the last wet spell ending date of the season. The second objective is for providing important input to decision-making support system in water-related sectors such as agriculture, hydrology, water resources and terrestrial and freshwater aquatic ecosystems, as well as diagnostics and prediction of rainfall. And, the third objective is to understand impact of global climate change on extreme rain events across the country.

2. Data used and division of the country for the present study

Daily rainfall data of the country on 1-degree latitude-longitude spatial resolution for

the period 1951-2007 developed by the India Metrological Department (IMD) (Rajeevan et. al., 2006) is utilized in the present study. In the past, due to large geographical area (~3.3 million km2), the country was divided into certain zones/subdivisions for different purposes, agriculture, hydrology, ecology, daily weather forecast, long range rainfall prediction, monitoring rainfall variations etc (Sontakke and Singh, 1996 and references therein). The division done for a particular purpose was not found suitable for another. A new division of the country has been done here for understanding climatology and variability of wet and dry spells across the country. Following physico-climatological factors are qualitatively considered in order to divide the country into an optimum number of subregions.

a) Topographic features; b) Spatial pattern of the mean annual/southwest monsoon rainfall; c) Physiographic characteristics, coast, plateau, plains, valley, desert, etc. (NATMO, 1986); d) Drainage pattern (NATMO, 1996); e) Normal onset and withdrawal dates of southwest monsoon across the country

(IMD, 1943); and f) Daily rainfall data available as 1-degree raster.

Based on topographic considerations the country is divided into seven regions e.g. West Coast, Central Highlands, Eastern Himalaya-North Eastern Range (northeast India or NEI; SR18), Western Himalaya-Punjab Plains (extreme northern India or ENI; SR19), peninsula (excluding West Coast), Indo-Gangetic Plains (IGPs) and northwest dry province. Considering mean annual/monsoon rainfall, the peninsula is again divided into central peninsula and east coast; and the West Coast into extreme southwest peninsula (ESWP; SR1),

4

 

southern central West Coast (SCWC; SR3), northern central West Coast (NCWC; SR5) and northern West Coast (NWC; SR8). By incorporating physiographic features the IGPs are further divided into 3 subregions, eastern IGP (EIGP; SR14), central IGP (CIGP; SR17) and western IGP (WIGP; SR16); central highlands into western central India (WCI; SR12) and eastern central India (ECI; SR13) and northwestern India into 2 subregions, northern northwest India (NNWI; SR15) and southern northwest India (SNWI; SR11). Inclusion of drainage pattern as a factor in the classification, the central peninsula and east coast are divided into 6 subregions, extreme southeast peninsula (ESEP; SR2), central southeast peninsula (CSEP; SR4), southern central peninsula (SCP; SR6), central east coast (CEC; SR7), northern central peninsula (NCP; SR9) and northern east coast (NEC; SR10). Daily rainfall data is available as 1-degree raster therefore; rectangular boxes are drawn to arrive at final classification of 19 subregions. While drawing boundaries it is checked that, the monsoon advanced northward with a distinct onset date from one subregion to another and also withdrew distinctly southward. The subregions are shown in Fig.1, and geographical area and some important climatological information about the subregions is given in Table 1. Geographical area of the subregions ESWP, SCWC, NCWC, NWC and NEI (SRs 1, 3, 5, 8 and 18) varies from 60,000 to 100,000 km2, that of NEC, WCI, NNWI, WIGP and ENI (SRs 10, 12, 15, 16 and 19) 200,000-300,000 km2 and that of other 9 subregions 100,000-200,000 km2. It is expected that characteristics of wet and dry spells on subregional scales would provide useful information for science (to understand physics and dynamics of rain-producing weather systems), services (operational weather forecasting) and society (agriculture, ecology, hydrology and water resources). Yearwise area-averaged daily rainfall data of the individual sub-regions has been prepared from simple arithmetic mean of the 1° lat-lon grids in the particular sub-region.

3. Objective criterion for identification of wet and dry spells

A dynamic criterion has been developed and applied to identify yearwise wet and dry spells over the 19 subregions. Rainfall threshold used in the criterion is derived from local rainfall climatology that is ‘daily mean rainfall (DMR) of the climatological (long term mean or normal) summer monsoon period over the area of interest’. Computational steps of the schemes are as follows:

i. Computation of daily rainfall climatology of the full calendar year. ii. Calculation of daily mean rainfall (DMR) of the normal summer monsoon period

(mm/day). iii. Normalization of year-wise daily rainfall amounts by dividing by the respective

DMR. iv. Application of the 9-point Gaussian low-pass filter with weights 0.244, ±0.201,

±0.117, ±0.047, ±0.013 on the normalized daily rainfall values. v. Identification of continuous period with normalized, smoothed daily rainfall values

equal to or greater than 1.0 as wet spell (WS) and less than 1.0 as dry spell (DS).

Thus, in a low frequency mode intra-annual rainfall variation the WS (DS) is identified as ‘continuous period with daily rainfall equal to or greater than (less than) daily mean rainfall (DMR) of climatological monsoon period over the area of interest’. Climatological onset and withdrawal dates of the southwest monsoon period over the sub-regions have been calculated (step no. ii) from yearly value of the respective parameters

5

 

manually picked up from the charts showing year-wise ‘Advance of Southwest Monsoon’ and ‘Withdrawal of Southwest Monsoon’ routinely prepared by the operational wing (Weather Central) of the India Meteorological Department (IMD), Pune. Yearwise charts of the ‘Advance of Southwest Monsoon’ are available for the period 1960-2007, and those of the ‘Withdrawal of Southwest Monsoon’ for the period 1975-2007. Climatological onset as well as withdrawal dates was found close to the normal onset and withdrawal dates shown on the earlier published charts of the IMD (IMD, 1943). For the 19 sub-regions, climatological onset and withdrawal dates and total monsoonal rainfall and daily mean rainfall of the monsoon period are given in Table 1. The length of the window-width for low-pass filtering (step no. iv) is decided after few trials, such as 3, 5, 7, 9 and 11-term running means, and low-pass binomial and Gaussian filters with the different window-widths. The filter with longer window-width (9-point) is chosen to suppress isolated rain-spell of one/two day(s) duration. In order to apply unified filter, the Gaussian low-pass filter was chosen giving satisfactory results for the 19 sub-regions. However, even after this careful choice rain-spell of one/two day(s) for few years of every sub-region could not be suppressed and they were retained as wet spell for detailed analyses. The criterion  is dynamic  in the sense the rainfall threshold used  is  changing  from  subregion  to  subregion. Further, the rainfall threshold is derived from the monsoon rainfall climatology of the particular subregion which provides a scientific basis for the criteria. The monsoon season is the most reliable period of rainfall activities under Indian conditions, therefore the rainfall threshold can be regarded robust. Identification of wet and dry spells is done with respect to local rainfall climatology therefore it is relative rather than absolute. For different subregions the DMR exceeds the rate of potential evapotranspiration (PE; Rao et al., 1971) except ESEP, CSEP and NNWI where PE exceeds DMR. Climatological and fluctuation features of the following parameters of actual wet and dry spells have been studied. 

i. Number of wet/dry spells ii. Total rainfall of the wet/dry spells iii. Total duration of the wet/dry spells iv. Starting date of first wet/dry spell v. Ending date of last wet/dry spell vi. Rainfall amount of individual wet/dry spell vii. Rainfall intensity of individual wet/day spell viii. Duration of individual wet/dry spell

Besides these, characteristics of important parameters (duration, rainfall amount, rainfall intensity and starting date) of the extremes of wet and dry spells in respect of rainfall amount, rainfall intensity and duration have been studied in order to understand if there is any change taking place in the hydrological cycle of the area. In total 40 parameters of the two attributes wet and dry spells of the rainfall time-distribution have been examined in this study. Description of spatial variation of total rainfall due to wet/dry spells, and rainfall amount and rainfall intensity of individual wet/dry spell is given in comparison with annual rainfall distribution across the country. Spatially-averaged sub-regional normal annual rainfall along the West Coast varies from 1795-3132mm; over central India 1077-1241mm; northeast 2089mm; extreme north 873mm; peninsula (excluding West Coast) 711-1424mm; Indo-Gangetic Plains 755-1423mm; and northwest dry province 324-617mm.

6

 

4. Results and Discussion

a. Climatological features of the wet spells

Normally, the ESEP (SR2) experiences the largest number (11) of WSs. The number decreases along the East Coast to 7 over EIGP (SR14), then decreases to 4 over northwestern India (SRs 11 and 15) from different directions (east, southeast and south). The standard deviation (SD) of inter-annual variation of number of WSs on sub-regional scale across India (hereafter will be referred to as ‘SDAI’) is about 2 (Table 2). Total duration of the WSs follows the similar pattern and decreases from 101 days (ESEP) to 29 days (NNWI) (SDAI = 14.3 days). Total rainfall due to WSs is however highest along the West Coast 1290-2130mm, followed by NEI 1262mm, and lowest over northwestern India 284mm. Over remaining parts the total rainfall varies from 495mm to 900mm following annual rainfall pattern with low values over central peninsula (NCP and SCP) and high values elsewhere. Features of individual WS are different along West Coast, duration varies from 10-12 days, rainfall amount 263-327mm and rainfall intensity 29-37mm/day; over other parts the duration varies from 7-10 days, rainfall amount 70-159 days and rainfall intensity 8-18mm/day. The SDAI of duration of individual WS is 3.1 days, rainfall amount 67.7mm and rainfall intensity 2.0mm/day. Interannual variation in rainfall intensity is a weak determinant of interannual variation of the WS rainfall amount (CC= ~0.45) and the respective annual rainfall (CC= ~0.36). Total duration of all the WSs is however strongly correlated with total rainfall. The correlation coefficient (CC) of total duration with total rainfall of the WSs is 0.72-0.98 across the 19 subregions and with the respective annual rainfall 0.70-0.93. The CC between total rainfall due to WSs and respective annual rainfall is 0.72-0.95.

Normally, the first WS starts the earliest on 19 March over the ESEP (SR2), followed by 23 April over NEI (SR18), 3 May over CSEP (SR4), 16 May over ENI (SR19), 21 May over ESWP (SR1), 23 May over SEC (SR7) and 30 May over SCP (SR6). Over peninsula, northeast and extreme north, thunderstorms associated first WS starts 11 to 55 days earlier than the onset of monsoon. Most parts of the country experience the first WS between 2 June and 23 June i.e. about 5 days earlier than the onset of monsoon. Over the NNWI the first WS starts on 22 June, 16 days earlier (8 July) than the monsoon onset. The SD of start of first WS is 8-11 days (smallest) over central and northern West Coast and 31-35 days over southeast peninsula and 30-53 days over extreme north/northwest. Hence, the time-invariant orographic effect reduces inter-annual variability of start of rains on windward side of large-scale moist air-flows from the surrounding oceans/seas. We believe that intense convection from elevated land mass and ample moisture supply by the perennial moist air stream from neighboring seas towards inland from beginning of the summer contribute to reducing inter-annual variation of the start of the first WS/season.

The cessation process of last WS starts from NNWI (SR15) on 12 September, the date progressively shifts towards east, southeast and south and end from the ESEP occurs on 16 December. In general, over most parts the last WS ends ~9 days earlier than the withdrawal of monsoon. However, from northeast, east coast and southeast peninsula the last WS ends after the withdrawal of monsoon: 4 days later from NEI, 7 days from NEC, 11 days from SCP, 22 days from SEC, 48 days from CSEP and 58 days from ESEP. This is because withdrawal of southwest monsoon is followed by onset of northeast monsoon over these subregions. The SD of ending date is 12-15 days (lowest) over southeast peninsula, 28 days along central WC and 31-35 days over northwest. Hence the time invariant orographic effect reduces variability of ending date of the rains on the windward side. The contribution of the

7

 

rainfall due to WSs to the respective annual total varies from 60 to 79% over different parts except over extreme north where it is only ~47% (Table 1). The last row of the table gives arithmetic mean of the sub-regions to provide an approximate value of the parameter for the whole country. The ESEP (SR2) experiences three WSs before onset of the summer monsoon and three WSs after the withdrawal; over CSEP (SR4) one WS before and two WSs after; over ESWP (SR1), SCP (SR6) and CEC (SR7) one WS before and one WS after; over NNWI (SR15) one WS before and no WS after; and over northeast (SR18) and extreme north (SR19) two WSs before and no WS after. Over remaining parts (SRs 3, 5, 8-14, 16 and 17) the first WS occurs with the onset of summer monsoon and the last WS ends with the withdrawal. Sometimes this large portion of the country is referred to as the core monsoon rainfall regime (CMRR; Ranade et al., 2008). b. Climatological features of the extremes of wet spells

Extreme wet spells are identified based on three conditions- extremely high rainfall amount, extremely high rainfall intensity and extremely long duration. Time series (1951-2007) of extreme wet spells for each of the three specified conditions has been prepared. Spatial features of duration, rainfall amount, rainfall intensity and starting date of different extreme WSs are described here in comparison with actual WS; climatology is given in Table 3,

1) EXTREME WET SPELL IN RESPECT OF RAINFALL AMOUNT Duration of extreme WSs in this category is longer (22-30 days) over south peninsula

and West Coast and shorter (13-21 days) elsewhere; rainfall amount higher (590-1051mm) along West Coast followed by 361-367mm over WCI, ECI and NEI and lower (154-338mm) elsewhere; and rainfall intensity greater (32-40mm/day) along West Coast followed by 17-20mm/day over east/northeast India and weaker (11-16mm/day) elsewhere. In general, the duration of extreme WSs is 2.16 times longer than that of the actual WS (SDAI = 8.7 days), rainfall amount 2.34 times (SDAI=173.2mm) and rainfall intensity 1.12 times (SDAI = 4mm/day). The spell occurs over parts of WC, northeast and extreme north during the month of June, over most parts during July and over southeastern peninsula during August, September and October from north to south. The SDAI is about 27.3 days which suggests that extreme WS can occur anywhere during the period of rainfall activities. The most extreme wet spell during 1951-2007 over different sub-regions is given in Table 4. The heaviest amongst sub-regional extremes occurred over the SCWC (SR3) which started on 28 June 1961, continued for 42 days and produced rainfall of 2080.2mm at the rate of 49.5mm/day. The year 1961 was most wet in the recorded instrumental data available from 1813.

2) EXTREME WET SPELL IN RESPECT OF RAINFALL INTENSITY Duration of such extreme WS is longer (15-23 days) over south peninsula and West

Coast and 6-12 days elsewhere; rainfall amount is higher (504-520mm) along West Coast and 259-292mm over ESEP and WCI and lower (87-205mm) elsewhere; and rainfall intensity greater (35-43mm/day) over West Coast followed by 20-25mm/day over east/northeast India and weaker (12-19mm/day) elsewhere. Broadly, duration of the extreme WS is 1.33 times (SDAI =8.5 days), rainfall amount 1.65 times (SDAI =185.6mm) and rainfall intensity 1.27 times (SDAI = 4.1mm/day) than that of actual WS. Rainfall amount and rainfall intensity is higher over high (annual) rainfall areas and lower over low rainfall areas. Duration however

8

 

shows typical spatial pattern- it increases from north/northeast to south/southeast. The extreme WS can start around 10 July along WC, 30 July over northwest, 10 August over east/northeast and 30 September-15 October over southeast peninsula. The SDAI is about 48.3 days which suggests occurrence of extreme WS anywhere between first and last WSs inclusive. The most intense extreme wet spell occurred over NWC (SR-8) starting on 20 June 2005 with rainfall amount 1042.2 mm, rainfall intensity 61.3mm/day and duration 17 days (Table 4).

3) EXTREME WET SPELL IN RESPECT OF DURATION Duration is longer (23-31 days) along West Coast and over central peninsula and

shorter (14-22 days) elsewhere; rainfall amount higher (576-1046mm) along West Coast followed by 341-366 days over central and northeast India and lower 145-314mm elsewhere; and rainfall intensity 30-39mm/day over West Coast and 9-19mm/day over other places. The average duration for the whole country is 2.23 times (SDAI =8.4 days), rainfall amount 2.28 times (SDAI =176.6mm) and rainfall intensity 1.04 times (SDAI =3.4mm/day) than that of actual WS. The duration increases from northwest to extreme southeast peninsula, however, rainfall amount and rainfall intensity is higher over highlands and low over plains. It starts around 30 June along West Coast and northeast, during July over areas north of 18˚N, mid-August over central peninsula (SRs 6 and 7) and 10-25 September over southeast peninsula. The SDAI is about 24.7 days suggesting occurrence of extreme WS anywhere during the period of seasonal rainfall activities. The longest wet spell on record occurred over SR2 starting on 28 June 2005 with duration 91 days, rainfall amount 1645.2 mm and rainfall intensity 18.1mm/day. In general, the different extreme wet spells are lengthier (2.16 times), wetter (2.09 times) and more intense (1.14 times) than the actual. c. Climatological features of the dry spells

Total duration of intervening DSs (3-10 in number) is least (45 days) over WCI and WIGP (SRs 12 and 16), increases to 59 days over NNWI, to 81 days over ENI, to 101 days over ENI and to 173 days over ESEP. At subregional scale across the country, the interannual variation (SD) in total duration of DSs is 17-61 days; the SD is larger where rainfall occurrences are spread over long period of the year and rainfall intensity is relatively low. Total rainfall due to DSs is high over higher (annual) rainfall areas and low over lower rainfall areas- over NEI 553mm, along West Coast 291-509mm and over CEC, NEC and EIGP 222-291mm and decreases to 28mm over NNWI. Duration of individual DS is least (11-12 days) over WCI, ECI, EIGP and CIGP (SRs 12, 13, 14 and 17), increases to 22 days over NNWI (SR15) and to 19 days over ESEP (Table 5). The SDAI is about 7.9 days suggesting large interannual variation in the duration on subregional scales. Mean rainfall amount of individual DS is 10mm over northwest, increases to 14mm over ESEP, to 37mm over ENI, to 70mm over NEI, and to 116mm along West Coast. The mean rainfall intensity during DS is 1-2mm/day over northwest and eastern south peninsula, 3-6mm/day over almost all northern India and 6-10mm/day along West Coast. The interannual variation in rainfall intensity during dry spells across the country is only about 1mm/day. The CC of the DS rainfall intensity with DS rainfall amount and annual rainfall is almost negligible. But total duration and total rainfall of the DSs are also the important contributors to the respective annual rainfall. The correlation (CC) between total duration and total rainfall of the DSs is in the range of 0.55 to 0.89, between total duration and respective annual rainfall 0.47-0.90, and between total rainfall and respective annual rainfall 0.47-0.97 across the 19 subregions. The DSs rainfall contributes about 17% to the respective annual total.

9

 

The first dry spell starts around 24 March over ESEP (SR2), 28 April over northeast (SR18), during May over ESWP (SR1), CSEP and CEC (SRs 4 and 7) and extreme north (SR19), and during June over most parts of the country except over WIGP (SR16) where it is 4 July. The SD of start of first DS is minimum (15-18 days) along West Coast and SNWI (SRs 3, 5, 8 and 11), large (32-35 days) along East Coast (SRs 2, 4 and 7) and over north/northwest (SRs 15 and 16), and largest (61 days) over extreme north (SR19). The last dry spell ends around 7-10 September over northwest (SRs 11 and 15) and the date progressively shifts to 6 December over the ESEP following normal withdrawal pattern of the summer monsoon. The SD of end of last DS is minimum (17-21 days) along south and central East Coast and maximum (28-35 days) along West Coast (SRs 1, 3 and 5) and central and northwestern India (SRs 11, 12 and 15). In general, the SD of start and end of DSs is lowest towards windward side of orographic barriers for large-scale moist monsoon flows and large elsewhere. The percentage contribution of rainfall due to DSs to respective annual total varies from 10% over NNWI (SR15) to 27% over NEI (SR18).

The CC between total duration of WSs and that of DSs is in the range -0.55 and 0.02 across the 19 subregions; it is mostly negative and weak. Similarly, the CC between total rainfall due to WSs and that due to DSs is mostly negative and weak in the range -0.40 to 0.02. The time-distribution of rainfall during a year into wet and dry spells over most parts of the country is such that its partitioning on the ground surface will produce more green water flow (actual and potential evapotranspiration) and green water storage (soil moisture) and less blue water flow (reservoir storage and water yield) and blue water storage (deep aquifer recharge). Under Indian condition partitioning of rainfall into green water and blue water is two-third and one-third respectively (Ranade et al., 2008). d. Climatological features of the extremes of dry spells Extreme dry spells are identified based on three conditions- extremely low rainfall amount, extremely low rainfall intensity and extremely long duration. Climatology of duration, rainfall amount, rainfall intensity and starting date of the extreme dry spell parameters is given in Table 6. A description of spatial features of the parameters of different extreme DSs is provided in comparison with actual dry spells (Table 5).

1) EXTREME DRY SPELL IN RESPECT OF RAINFALL AMOUNT Duration is 11-16 days over northern India and 4-7 days over all other subregions;

rainfall amount 24-49mm over West Coast, 1-7mm over southern peninsula and 9-17mm elsewhere; and rainfall intensity near-zero (0.4-2mm/day) over hot, dry regions of northwest and southern peninsula, 6-11mm/day along West Coast and 3-4mm/day over other places. Broadly, the duration of extreme DSs is shorter (0.42 times) than the actual DS; the rainfall amount lower (0.34 times) and the rainfall intensity slight weaker (0.96 times). Such extreme DS occurs during July over western India (SRs 1, 3, 5, 8, 11 and 15) and northeastern parts, during August over central and eastern parts, and during 8-11 September over southeastern peninsula (SRs 2 and 4). The most extreme DS of nil rainfall occurred over dry regions of ESEP during 24-25 December 1965, SCP on 13 September 1960 and NNWI on 31 July 1983 (Table 7).

2) EXTREME DRY SPELL IN RESPECT OF RAINFALL INTENSITY Duration of extreme DSs is shortest (10 days) over CIGP (SR17) and it increases in

all directions: 37 days over ENI, 26 days over NNWI, 23 days SCWC and 13 days ESEP; rainfall amount is 7mm over NNWI, 3-9mm over SCP and SEP, 72-154 along West Coast

10

 

and 18-51mm elsewhere; and rainfall intensity near-zero (< 2mm/day) over NNWI, SCP, CEC, CSEP and ESEP, 3-8mm/day along WC and 2-3mm/day elsewhere. For the country as a whole, the duration of extreme DS is 1.28 times than that of actual DS; rainfall amount 0.87 times and rainfall intensity 0.64 times. Such extreme DS occurs around 18 June over ENI (SR19), during July along the West Coast and SNWI (SRs 1, 3, 5 and 11), CIGP (SR17) and NEI (SR18), during August over almost entire central India and ESEP (SR2) and around 21 September over CSEP (SR4). The most extreme DS of lowest rainfall intensity (0.0mm/day) occurred over ESEP during 24-25 December 1965 and over NNWI on 31 July 1983 (Table 7).

3) EXTREME DRY SPELL IN RESPECT OF DURATION Duration is shortest (24 days) over CIGP and ECI and it increases to 32 days over

NEI, 36 days over NNWI, 50 days over ENI and 58 days over ESEP; rainfall amount is lowest (13mm) over NNWI and it increases to 220mm along West Coast, 89mm over NEC, 150mm over ENI and 78mm over ENI; and rainfall intensity near-zero (<2mm/day) over NNWI and southeast peninsula, 5-9mm/day along West Coast and 2-5mm/day elsewhere. Broadly, the duration of extreme DSs is 1.28 times the duration of actual DS, rainfall amount 1.93 times and rainfall intensity 0.89 times. It occurs around 15 May over ESEP (SR2), during June over CSEP (SR4) and ENI (SR19) and during July-August over remaining parts. The most extreme dry spell occurred over NNWI (SR15) starting from 8 July 1980; duration was 168 days, rainfall amount 56.3 mm and rainfall intensity 0.3mm/day (Table 7).

To comprehend broad spatial climatological features of the dry and wet spells across the country, a generalized description of the parameters in five categories is provided.

• Number and total duration of wet/dry spells and duration of actual and different extreme wet/dry spells- highest value in the extreme southeast peninsula and decreases towards north/northwest as the tropical and oceanic influences decrease.

• Total rainfall of wet/dry spells and rainfall amount and rainfall intensity of actual and extreme wet/dry spells- high value in the orographic regions (West Coast and central, northeast and extreme north India, and lowest over plains (peninsula, Indo-Gangetic Plains and northwest dry provinces.

• Start of the first wet spell- earliest in extreme southeast peninsula and latest in northwestern India.

• End of the last wet spell- earliest from northwestern India and latest from extreme southeast peninsula.

• Occurrence of different extreme wet/dry spells- during June through August over most parts of northern India and during September through December over south peninsula.

Interesting to note is that normally 68% of the annual rainfall occurs during WSs and

17% during DSs (intervening period of subdued and/or no rainfall activities) identified by the present objective criteria. So, 85% of the annual rainfall occurs between first WS and last WS and only 15% as isolated random convection spread over remaining parts of the year. It is close to widely known rainfall climatology that about 80% of the annual rainfall over the country is seasonal. Spatial variation of this seasonal contribution over most subregions is between ~80% and ~90%, which is relatively small. Hence, identification of the wet and dry spells can be regarded robust. To get these robust results it is essential to use variable rainfall threshold for different subregions. Arbitrary, fixed choice of the threshold and its uniform application (Ananthakrishanan and Soman, 1988; Ananthakrishanan et al., 1967; Aviad et al.,

11

 

2004; Bai et al., 2007; Cook and Heerdegen, 2001; Dash et al., 2009; Deni et al., 2008; Raman, 1974; Stern, 1981) will produce unrealistic results for the subregions. If rainfall threshold is larger than the daily mean rainfall (DMR) of the subregion, in low rainfall areas there will be less and shorter WSs, large and longer DSs, some or most years without any WS, late start of first WS and early end of last WS while opposite will be the features of rainfall occurrences in high rainfall areas. And if rainfall threshold is smaller than the DMR, in high rainfall areas there will be large and longer WSs, less and shorter DSs, some or most years without any DS, early start of first WS and late end of last WS while opposite will be the rainfall features in low rainfall areas. e. Fluctuation features of the wet and dry spells and their extremes

In this section fluctuations of wet and dry spells and their extremes across the country are reported. Each of the 760 time series (40 parameters multiplied by 19 subregions) has been examined at the following 3 levels. i. Visual examination of time series plots of the actual and 5-point binomial low-pass

filtered values; ii. Fitting least-squares linear trend; and iii. Student’s t-test for difference between means of two sub-periods of the available data.

From visual examination each of the 40 parameters is found exhibiting wide range of fluctuation characteristics across the country such as short-term/long-term rising/falling trend, stationary etc.– the time series of selected parameters are displayed in Figures 21-32 in order: i) number of WSs; ii) mean rainfall intensity during WS; iii) mean duration of WS; iv) start of first WS; v) rainfall amount of extreme WS in respect of (i.r.o.) rainfall amount; vi) rainfall intensity of extreme WS i.r.o. rainfall intensity; vii)duration of extreme WS i.r.o. duration; viii) mean rainfall intensity during DS; ix) mean duration of DS; x) rainfall amount of extreme DS i.r.o. rainfall amount; xi) rainfall intensity of extreme DS i.r.o. rainfall intensity and xii) duration of extreme DS i.r.o. duration. Least-squares linear trend has been fitted to each series to know broad nature (+ve/-ve) of long term trend and an abstraction of results is as: 364 series show increasing tendency (29 of which are significant at 5% level and above) and 396 decreasing tendency (42 of which are significant at 5% level and above). To know if there is any change in the mean of recent 30-year period (1978-2007) compared to previous 27 years (1951-1977) Student’s t-test has been applied and the results are: 341 series show increase in the recent 30-year (63 of which are significant at 5% level and above) and 419 show decrease (31 of which are significant at 5% level and above). The exercise is repeated and the mean of recent 20-year period 1988-2007 is compared with the mean of earlier period 37 years (1951-1987) and the results are: 353 series show positive tendency (33 of which are significant at 5% level and above) and 407 show decrease (12 of which are significant at 5% level and above). The detailed results for wet spells are given in Table 8 and presented in Fig. 33, and dry spells in Table 9 and Fig. 34 respectively. Further the mean of recent 10-year period (1998-2007) is compared with previous 47 years (1951-1997) and the results are: 361 show increases (28 of which are significant at 5% level and above) and 399 show decrease (8 of which are significant at 5% level and above). Spatially-coherent robust long-term trend is not seen in any of the parameters during 1951-2007. However, persistent positive/negative tendencies spread over 11 or more subregions are seen in some of the parameters. Notable among them are decreasing tendency in duration and increasing tendency in rainfall intensity of actual and extreme WSs, and increasing tendency in duration

12

 

and decreasing tendency in rainfall intensity of actual and extreme DSs. These changes in rainfall time distribution appear to be related to changes in frequency and duration of low pressure systems; low pressure areas (LPA) and depressions/storms (DDS). Though the number and duration of LPA shows significant increasing trend at the rate of 1.4/10-year and 8.1 days/10-year respectively, the number and duration of DDS shows significant decreasing trend at the rate of 1.5/10-year and 6 days/10-year respectively (Jadhav and Munot, 2008). Further it appears there is some change in middle tropospheric (850-500 hPa) circulations (trough, convergence zone and cyclonic circulations) due to warmer environment. In random occurrences process if surface and middle tropospheric circulations were coherent shorter, intense WSs occurred otherwise longer, severe DSs occurred. But this proposition required investigation which can be done in separate study. Examining association between features of wet spells (frequency, intensity and duration) and that of rain-producing weather systems (thunderstorms, vortices, line systems and wave disturbances) will be topic of our future research. Compared to 1951-1987, during 1988-2007 the first WS started about 6 days earlier and last WS ended about 2 days earlier. Over most parts of the country the first WS starts during May-June and the last WS ends during September-October. It may be noted that the July, August and September rainfalls show declining tendency in recent years/decades over the country. But the April May and June rainfalls show rising tendency. Therefore, the first WS is starting slightly earlier (~6 days). Similarly, because of rising tendency in October rainfall and declining tendency in November, December and January rainfall the last WS is ending ~2 days earlier. Consequently, the period between start of first WS and end of last WS is ~4 days longer in recent years. Following two additional tests have also been applied to see if the above results are robust: i. Mann-Kendall Rank test of randomness against trend (WMO, 1966); and ii. Wilcox-Mann-Whitney rank-sum test for two independent samples (Wilks, 2006). These tests produced broadly the same results as obtained by the two earlier tests i.e. least-squares linear trend and Student’s t-test. Hence the results reported in this paper are robust. f. Limitations Rainfall threshold derived from local climatology is used to understand broad climatology of WSs and DSs and their climatic and hydroclimatic variability, however, depending upon practical problems (agriculture, hydrology etc.) and meteorological conditions (evaporation/evapotranspiration) regional threshold independent of present dynamic framework can be adopted. But its application will be limited to the particular region. Even in CMRR (between parallels 18°N and 32°N; and between meridians 68°E and 88°E) standard deviation of onset and withdrawal date is ~7 days (5-9 days), but standard deviation of start of first WS is ~21days and end of last WS ~23 days. This suggests that for determining onset and withdrawal dates, rainfall alone is not sufficient. Other parameters like meteorological conditions over the particular area (geopotential height of selected isobaric levels, precipitable water, cloud amount, OLR etc.), strength of large-scale monsoon flow (meridional wind along Somali Coast, zonal wind over central Arabian Sea, meridional wind over Bay of Bengal, resultant wind over South China Sea etc.), intensity of subtropical high/anticyclone (Tibetan Anticyclone, Northwest Pacific subtropical High, Mascarene High and Australian High) and state of Tropical Easterly Jet (TEJ) should also be considered. This

13

 

is being done as a separate study. In the subregional rainfall analysis the station rainfall amount is highly suppressed. Therefore, subregional extreme rainfall cannot be used for estimation of probable maximum precipitation (PMP) and standard project storm (SPS). Using longest instrumental monthly/seasonal/annual rainfall Sontakke et al. (2008a) reported that mean annual rainfall over the country during 1965-2006 was less by 4.23% compared to 1931-1964. Most of the daily rainfall data (1951-2007) used in the present analysis is from the dry epoch (1965-2007). Therefore, if 20-30 years of data prior to 1951is included in the present analysis the results may be somewhat different. 5. Summary and conclusions

A summary of the main results is as follows: 1) Normally, number of wet spells (WSs) across the country varies from 4 (NWI) to 11

(ESEP). The total duration of the WSs varies from 29.2 days (NWI) to 100.4 days (ESEP) and the total rainfall from 283.7 mm (NWI) to 2130.4 mm (SCWC). The duration of individual WS varies from 6.9 days (NEC) to 11.5 days (SCWC), rainfall amount 69.7mm (SCP) to 426.1mm (SCWC) and rainfall intensity 7.6mm/day (ESEP) to 36.9mm/day (SCWC).

2) The total duration of the intervening DSs varies from 44.5 days (WIGP) to 172.5 days (ESEP) and the total rainfall from 27.5 mm (NWI) to 509.1 mm (ESWP). The duration of individual DS varies from 10.3 days (SCIGP) 19 days (NWI), the rainfall amount from 8.9 mm (NWI) to 114.9 mm (SCWC) and the rainfall intensity 0.6 mm/day (NWI) to 9.8 mm/day (SCWC).

3) The country as a whole gets more than 6 WSs during 31 May through 8 October each of duration 8.8 days and rainfall amount 152.8mm. The duration of the intervening DS is 14 days and the rainfall amount 45.3mm. The total rainfall and the total duration of the WSs (DSs) and the respective annual rainfall are highly correlated; the CC between total rainfall and total duration is ~0.91 (~0.77), between total rainfall and annual rainfall ~0.87 (~0.86) and between total duration and annual rainfall ~0.84 (~0.70). The WSs contribute ~68% and the DSs ~17% to the respective annual rainfall.

4) The first WS starts with the onset of the summer monsoon, and the last WS ends with the withdrawal across the country except over south/southeast peninsula, northeast, northwest and extreme north where wet spells are caused by extra-monsoonal systems (thunderstorms, easterly/westerly waves, tropical cyclones and northeast monsoon) that occur before and after the summer monsoon period. Interannual variation of start and end of the wet spells is least on the windward side of the orographic barriers (West Coast and East Coast).

5) There is a weak tendency (statistically not significant) for the first WS to start ~6 days earlier and the last WS to end ~2 days earlier; consequently period of rainfall activities is ~4 days longer in recent years.

6) The different extreme WSs (in respect of rainfall amount, rainfall intensity as well as duration) are lengthier (1.9 times), wetter (2.1 times) and more intense (1.1 times) than the actual WS.

7) The extremely DSs in respect of rainfall amount are shorter (0.42 times), drier (0.34 times) and slightly weaker (0.96) than the actual DS; in respect of rainfall intensity they are longer (1.28 times) but drier (0.87 times) and weaker (0.64 times); and in respect of duration they are much longer (2.26 times) and wetter (1.93 times) but somewhat less intense (0.89 times).

14

 

8) The number, total rainfall and total duration of the WSs and the DSs show increasing tendency in recent years/decades over majority of the sub-regions. However, fluctuation of individual WS and DS is contrasting. The actual and extreme (in respect of rainfall amount and rainfall intensity) WSs are slightly shorter and rainfall intensity higher while duration of actual and different extreme DSs are slightly longer and rainfall intensity lower.

9) Spatially coherent significant long term trend is not seen in any of the 40 WS/DS parameters over the period 1951-2007 examined in the present study.

In the tropical monsoon environment of India seasonally occurring rainfall with

relatively shorter WSs (~9 days) and longer DSs (~14 days) is more favourable for agricultural and ecological activities than hydrological and water resources purposes. The characteristics of wet and dry spells across India provide some evidence to the hypothesis that ‘hydrological cycle is supposed to be intensified under global climate change background’ (IPCC, 1996). Though lower troposphere (surface-600 hPa) shows warming trend across the globe, the upper troposphere (400- 150 hPa) shows cooling trend resulting into weaker Tibetan Anticyclone and weaker southwest monsoon flow and a diminuendo phase in rainfall over India (Sontakke et al. 2008b). Acknowledgements. The authors are extremely grateful to Prof. B.N. Goswami, Director, Indian Institute of Tropical Meteorology, Pune (INDIA) for necessary facilities to pursue this study, and to Dr. (Mrs) Ashwini A. Kulkarni for critically reviewing the manuscript. The 1-degree raster data of daily rainfall of the period 1951-2007 for the country was provided by the India Meteorological Department, Pune which is thankfully acknowledged

15

 

REFERENCES

Ananthakrishnan, R., U.R. Acharya, and A.R. Ramakrishnan, 1967: India Meteorological Department (IMD), Forecasting Manual, Part IV, 18.1, Monsoons of India, p.9.

Ananthakrishnan, R., and M.K Soman, 1988: The onset of the southwest monsoon over Kerala: 1901-1980. Journal of Climatology, 8, 283-296.

Aviad, Y., H. Kutiel, and H. Lavee, 2004: Analysis of beginning, end, and length of the rainy season along a Mediterranean-arid climate transect for geomorphic purposes. J. Arid Environments, 59, 189-204.doi:10.1016/j.jaridenv.2004.01.013.

Bai, A., P. Zhai, and X. Liu, 2007: Climatology and trends of wet spells in China, Theoretical Applied Climatology, 88, 139-148.

Byun, H.R., and D.K. Lee, 2002: Defining three rainy seasons and the hydrological summer monsoon in Korea using Available Water Resources Index, J. Meteor. Soc. Japan, 80, 33-44.

Byun, H.R., and D.A. Wilhite, 1999: Objective quantification of drought severity and duration, J. Climate, 12, 2747-2756.

Cook, G.D., and R.G. Heerdegen, 2001: Spatial variation in the duration of the rainy season in monsoonal Australia. Int. J. Climatol., 21, 1723-1732.

Dai, A, K.E. Trenberth, and T.R. Karl, 1998: Global variations in droughts and wet spells: 1900-1995, Geophy. Res. Lett., 25, 17, 3367-3370.

Dash, S.K., M.A. Kulkarni, U.C. Mohanty, and K. Prasad, 2009: Changes in the characteristics of rain events in India, J. Geophy. Res., 114, D10109, doi:10.1029/2008JD010572, 2009.

Deni, S. M., A. A. Jemain, and K. Ibrahim, 2008: The spatial distribution of wet and dry spells over Peninsular Malaysia, Theor. Appl. Climatol., DOI 10.1007/s00704-007-0355-8.

India Meteorological Department (IMD), 1943: Climatological Atlas for Airmen.3, India Meteorological Department, New Delhi, 100 plates.

Jadhav, S. K., and A.A. Munot, 2008: Warming SST of Bay of Bengal and decrese in formation of cyclonic disturbances over the Indian region during southwest monsoon season. Theor. Appl. Climatol., DOI 10.1007/s00704-008-0043-3.

NATMO, 1986: National Atlas of India- Physiographic regions of India, Third edition, National Atlas and Thematic Mapping Organisation (NATMO), Kolkata, India, Plate No. 41.

NATMO, 1996: Land Resource Atlas- Drainage, National Atlas and Thematic Mapping Organisation (NATMO), Kolkata, India, Plate No.3.

Rajeevan, M., Bhate J, Kale J. and Lal B., 2006: High resolution daily gridded rainfall data for the India region : Analysis of break and active monsoon spells. Current Science, 91, 296-306.

16

 

Raman, C.V.R., 1974: Analysis of commencement of Monsoon rains over Maharashtra State for agricultural planning, Scientific Report 216, India Meteorological Department, Poona, 23 pp.

Ranade, Ashwini, N. Singh, H.N. Singh, and N.A. Sontakke, 2008: On variability of hydrological wet season, seasonal rainfall and rainwater potential of the river basins of India (1813-2006). J. Hydrol. Res. Devpt., 23, 79-108.

Rao, K.N., George, C.J. and Ramasastri, K.S. (1971) Potential evapotranspiration (PE) over India. Pre-published Sci. Rep. No. 136. Meteorological Office: Poona.

Singh, N., 1986: On the duration of the rainy season over different parts of India. Theoretical and Applied Climatology, 37, 51-62.

Sontakke, N.A., H.N. Singh, and N. Singh, 2008a: Chief features of physiographic rainfall variations across India during instrumental period (1813-2006). Contribution from IITM, Research Report No. RR-121, Indian Institute of Tropical Meteorology, Pune, India, 128 pp.

Sontakke, N.A. and Singh, N. (1996) Longest instrumental regional and all-India summer monsoon rainfall series using optimum observations: reconstruction and update. The Holocene, 6, 315-331.

Sontakke, N.A., N. Singh, and H.N. Singh, 2008b: Instrumental period rainfall series of the Indian region (1813-2005): revised reconstruction, update and analysis. The Holocene, 17, 1055-1066.

Stern, R.D., M.D. Dennett, and D.J. Garbutt, 1981: The start of the rains in West Africa. J. Climatol, 1, 59-68.

Wilks, D.S., 2006: Statistical Methods in the Atmospheric Sciences, 2nd Edn., Elsevier, Amsterdam, 627 pp.

WMO, 1966: Climatic Change, World Meteorological Organization, Technical Note 79, WMO No.195-TP-100, Geneva. 79 pp.

17

 

TABLE 1. Some important information about the 19 sub-regions: geographical area, mean annual rainfall, mean onset and withdrawal dates of the summer monsoon, total monsoonal rainfall and daily mean rainfall as well as daily mean potential evapotranspiration (PE) of the monsoon period.

Sub-

region Name of the subregion

Approximate Area

(sq. km)

Normal annual rainfall (mm)

Onset of SW monsoon

(SD in days)

Withdrawal of SW

monsoon (SD in days)

Rainfall of monsoon

period (mm)

Daily mean

rainfall of monsoon

period (mm/day)

SR 1 Extreme southwest peninsula (ESWP) 60,000 2700.8 1 Jun (7) 19 Oct (6) 2061.2 14.6SR 2 Extreme southeast peninsula (ESEP) 110,000 962.0 2 Jun (7) 19 Oct (6) 365.6 2.6SR 3 Southern central West Coast (SCWC) 60,000 3131.9 5 Jun (7) 17 Oct (6) 2724.4 20.2SR 4 Central southeast peninsula (CSEP) 150,000 859.4 6 Jun (7) 17 Oct (6) 464.4 3.5

SR 5 Northern central West Coast (NCWC) 60,000 2220.1 9 Jun (7) 13 Oct (5) 1990.3 15.7

SR 6 Southern central peninsula (SCP) 120,000 710.5 11 Jun (6) 14 Oct (6) 538.1 4.3SR 7 Central East Coast (CEC) 110,000 1025.8 12 Jun (6) 15 Oct (7) 715.5 5.7SR 8 Northern West Coast (NWC) 90,000 1794.5 13 Jun (6) 7 Oct (8) 1648.0 14.1SR 9 Northern central peninsula (NCP) 180,000 1012.7 15 Jun (6) 8 Oct (7) 855.7 7.4SR 10 Northern East Coast (NEC) 210,000 1423.9 16 Jun (6) 12 Oct (6) 1080.5 9.1SR 11 Southern northwest India (SNWI) 190,000 616.9 23 Jun (9) 28 Sep (6) 534.7 5.5SR 12 Western central India (WCI) 240,000 1077.2 21 Jun (8) 3 Oct (6) 928.3 8.8SR 13 Eastern central India (ECI) 150,000 1240.7 20 Jun (7) 7 Oct (6) 1036.2 9.4SR 14 Eastern Indo-Gangetic Plain (EIGP) 120,000 1422.9 13 Jun (6) 10 Oct (6) 1145.5 9.5SR 15 Northern northwest India (NNWI) 280,000 323.9 8 Jul (8) 19 Sep (8) 230.4 3.1SR 16 Western Indo-Gangetic Plains WIGP) 240,000 755.4 30 Jun (8) 27 Sep (8) 598.2 6.6 SR17 Central Indo-Gangetic Plain (CIGP) 150,000 1140.2 21 Jun (7) 3 Oct (6) 918.8 8.8SR 18 Northeast India (NEI) 60,000 2089.2 6 Jun (6) 15 Oct (5) 1461.6 11.1SR 19 Extreme northern India (ENI) 290,000 873.3 1 Jul (8) 23 Sep (9) 544.6 6.4

18

 

TABLE 2. Mean and standard deviation (SD) of 9 parameters of the wet spells for the 19 sub-regions.

Sub regions

WS (SD)

TRAINWS

(SD) in mm TDUR

WS

(SD) in days RI

WS

(SD) in mmRAIN

WS

(SD) in mm DUR

WS

(SD) in daysSTART

WS

(SD in days) END

WS

(SD in days) age

WS%

(SD)

ESWP 6.9 (2.0) 1817.7 (538.1) 65.5 (17.3) 27.6 (2.5) 263.4 (139.1) 9.5 (4.3) 21 May (17.5) 22 Oct (25.8) 66.1 (8.6)

ESEP 10.7 (2.1) 778.8 (275.5) 100.4 (20.0) 7.6 (1.5) 72.8 (35.8) 9.4 (2.5) 19 Mar (32.6) 16 Dec (11.6) 79.3(8.4)

SCWC 5.0 (1.4) 2130.4 (503.1) 57.7 (12.7) 36.9 (3.5) 426.1 (201.6) 11.5 (5.1) 2 Jun (11.1) 17 Sep (27.7) 67.3 (7.9)

CSEP 9.0 (2.2) 663.0 (216.9) 80.5 (17.6) 8.2 (1.5) 73.7 (37.0) 8.9 (2.6) 3 May (31.0) 4 Dec (15.1) 75.92(7.1)

NCWC 4.8 (1.2) 1567.3 (438.9) 51.8 (12.2) 30.1 (3.3) 326.5 (134.6) 10.9 (3.7) 7 Jun (8.7) 19 Sep (27.9) 69.6 (7.8)

SCP 7.3 (1.9) 508. 9 (168.2) 57.6 (17.7) 8.9 (1.0) 69.7 (31.1) 7.9 (2.9) 30 May (26.1) 25 Oct (20.5) 70.1 (9.2)

CEC 8.3 (1.9) 711.2 (202.0) 61.2 (15.5) 11.6 (1.6) 85.7(25.3) 7.4 (2.0) 23 May (34.6) 7 Nov (19.1) 68.3 (8.4)

NWC 4.7 (1.3) 1289.9 (386.3) 45.1 (11.8) 28.7 (3.6) 280.4 (98.9) 9.7 (3.0) 15 Jun (8.1) 15 Sep (18.7) 70.7 (8.4)

NCP 6.1 (1.7) 707.8 (216.8) 52.2 (14.6) 13.5 (1.5) 116.0 (59.2) 8.6 (3.9) 9 Jun (27.4) 6 Oct (24.4) 68.5 (11.0)

NEC 8.1 (1.7) 900.2 (235.9) 55.5 (13.8) 16.2 (1.3) 111.1 (31.2) 6.9 (1.8) 8 Jun (24.4) 19 Oct (21.4) 62.3 (9.1)

SNWI 4.1 (1.5) 494.6 (225.3) 35.9 (14.0) 13.5 (2.5) 120.6 (56.5) 8.8 (3.5) 21 Jun (15.8) 16 Sep (30.6) 77.2 (12.7)

WCI 4.8(1.3) 761.0 (194.3) 48.4 (12.4) 15.8 (1.6) 158.5 (69.0) 10.1 (4.3) 21 Jun (18.1) 21 Sep (28.2) 69.8 (8.1)

ECI 5.6(1.6) 820.0 (203.8) 51.4 (11.9) 15.9 (1.1) 146.4 (63.8) 9.2(3.8) 18 Jun (27.2) 24 Sep (18.6) 65.4 (7.8)

EIGP 6.8 (1.6) 879.9 (272.0) 52.1 (14.1) 16.8 (1.5) 129.4 (38.0) 7.7 (2.0) 7 Jun (20.5) 1 Oct (14.4) 60.6(10.3)

NNWI 3.8 (1.5) 283.7 (173.7) 29.2 (12.8) 9.2 (3.1) 78.8 (53.2) 7.7 (4.2) 22 Jun (30.4) 12 Sep (34.9) 77.4 (14.8)

WIGP 4.7(1.5) 531.2 (160.1) 44.5 (12.9) 11.9 (1.0) 113.0 (46.2) 9.4 (3.7) 23 Jun (30.5) 19 Sep (19.9) 68.9 (10.9)

CIGP 6.5 (1.5) 776.6 (219.1) 45.8 (10.7) 16.8(1.9) 119.5 (35.7) 7.1 (1.9) 14 Jun (28.8) 25 Sep (14.4) 66.9 (8.7)

NEI 9.2 (2.2) 1261.6 (319.2) 69.2 (15.8) 18.2 (1.5) 137.1 (36.4) 7.5 (1.8) 23 Apr (25.0) 9 Oct (22.6) 59.5 (8.7)

ENI 5.7(1.6) 486.3 (166.7) 46.9 (14.5) 10.3 (0.9) 74.9 (28.9) 8.2 (2.4) 16 May (52.9) 20 Sep (23.1) 47.3 (11.5)

INDIA 6.4 (1.7) 914.2 (269.3) 55.0 (14.3) 16.7 (1.9) 152.8 (64.3) 8.8 (3.1) 31 May (24.8) 8 Oct (22.0) 68.0 (9.4)

WS indicates mean number, TRAINWS mean total rainfall and TDUR

WS mean total duration of the WSs;

RIWS mean rainfall intensity, RAIN

WS mean rainfall amount and DURWS mean duration of the individual

WS; STARTWS mean starting date of the first WS; END

WS mean ending date of the last WS; and ageWS

%

mean percentage contribution of the WSs rainfall to the annual total.

19

 

TABLE 3. Mean and standard deviation (SD) of the parameters of the extreme wet spells in respect of rainfall amount, rainfall intensity and duration.

Sub region

RAINSW~ RISW

~ DURSW~

RAINRAINSW

~

(SD) in mm

RAINRISW

~

(SD) in mm/day

RAINDURSW

~

(SD) in days

RAINSTARTSW

~

(SD in days)

RIRAINSW

~

(SD) in mm

RIRISW

~

(SD) in mm/day

RIDURSW

~

(SD) in days

RISTARTSW

~

(SD in days)

DURRAINSW

~

(SD) in mm

DURRISW

~

(SD) in mm/day

DURDURSW

~

(SD) in days

DURSTARTSW

~

(SD in days)

ESWP 716.7 (351.6) 31.7 (4.9) 22.5 (10.3) 27 Jun (23.6) 504.1 (369.4) 35.9 (6.3) 14.4 (10.4) 10 Jul (53.4) 699.8 (361.7) 29.6 (4.6) 23.1 (10.0) 25 Jun (21.0)

ESEP 337.7 (229.6) 11.1 (3.8) 30.2 (14.8) 6 Oct (52.0) 291.6 (243.8) 12.3 (3.7) 23.3 (16.6) 16 Oct (70.4) 313.9 (236.2) 9.4 (3.2) 31.2 (14.1) 25 Sep (48.5)

SCWC 1051.5 (441.5) 39.8 (5.7) 26.4 (10.3) 26 Jun (16.2) 770.2 (535.6) 43.3 (5.5) 17.7 (12.0) 9 Jul (38.6) 1045.9 (447.6) 39.1 (5.7) 26.5 (10.1) 25 Jun (16.6)

CSEP 238.8 (193.3) 10.9 (3.5) 22.4 (12.2) 24 Sep (42.8) 188.4 (207.1) 12.5 (3.3) 15.0 (13.4) 2 Oct (62.9) 219.3 (199.3) 8.6 (2.5) 23.8 (11.5) 10 Sep (37.9)

NCWC 767.6 (297.6) 33.0 (4.7) 23.3 (8.3) 2 Jul (16.2) 577.8 (341.3) 35.3 (4.9) 16.5 (9.7) 9 Jul (37.6) 758.9 (305.3) 32.0 (4.8) 23.4 (8.1) 30 Jun (17.2)

SCP 187.7 (93.9) 10.5 (2.6) 18.3 (8.9) 22 Aug (33.8) 142.5 (111.2) 12.2 (2.6) 12.1 (9.6) 27 Aug (53.0) 180.9 (97.8) 9.4 (1.9) 18.8 (8.7) 14 Aug (33.7)

CEC 220.5 (76.9) 15.8 (6.4) 15.2 (6.0) 30 Aug (45.8) 174.8 (91.7) 18.1 (5.5) 9.8 (4.8) 27 Aug (62.3) 186.0 (79.4) 11.3 (3.3) 16.4 ( 5.5) 13 Aug (34.8)

NWC 589.7 (205.8) 32.6 (6.5) 18.3 (6.5) 11 Jul (17.4) 520.3 (259.9) 34.7 (6.7) 15.2 (8.2) 12 Jul (29.4) 576.0 (209.1) 30.9 (5.4) 18.6 (6.3) 11 Jul (17.1)

NCP 284.8 (119.7) 14.6 (2.8) 19.6 (7.6) 22 Jul (17.7) 205.0 (141.5) 16.7 (2.8) 12.3 (8.7) 28 Jul (42.6) 281.1 (118.8) 14.1 (2.8) 19.7 (7.4) 22 Jul (16.5)

NEC 270.0( 86.9) 17.6 (4.0) 15.8 (5.5) 30 Jul (28.8) 158.1 (102.9) 22.5 (4.5) 7.3 (5.2) 28 Aug (53.3) 263.5 (89.7) 16.4 (3.0) 16.2 (5.3) 28 Jul (24.0)

SNWI 268.6 (155.3) 15.7 (4.2) 16.9 (8.4) 16 Jul (23.3) 247.5 (166.4) 16.6 (3.9) 14.4 (9.1) 27 Jul (37.9) 263.7 (158.9) 14.8 (4.1) 17.1 (8.2) 15 Jul (22.1)

WCI 367.2 (135.1) 17.3 (2.7) 21.5 (8.4) 23 Jul (18.0) 258.7 (174.0) 18.7 (2.3) 13.9 (9.8) 2 Aug (39.7) 365.5 (135.2) 17.1 (2.7) 21.6 (8.3) 21 Jul (18.0)

ECI 344.7 (170.1) 17.03 (2.5) 20.5 (10.1) 17 Jul (23.9) 176.0 (131.2) 19.6 (3.0) 9.3 (7.5) 10 Aug (44.1) 341.2 (172.5) 16.4 (2.3) 20.8 (10.0) 17 Jul (22.1)

EIGP 295.3 (124.0) 18.4 (4.3) 16.4 (6.8) 18 Jul (28.4) 184.1 (105.4) 22.2 (4.4) 8.3 (4.7) 17 Aug (38.6) 290.5 (127.2) 17.4 (3.5) 16.7 (6.7) 13 Jul (24.5)

NNWI 154.4 (115.9) 11.4 (5.3) 12.7 (6.5) 27 Jul (30.0) 143.1 (119.5) 12.2 (5.4) 10.9 (6.9) 28 Jul (37.2) 144.6 (112.0) 10.2 (4.9) 13.0 (6.5) 19 Jul (24.7)

WIGP 253.7 (108.5) 12.4 (1.7) 20.7 (8.9) 25 Jul (17.2) 136.1 (114.4) 14.3 (1.9) 9.6 (8.5) 13 Aug (50.5) 252.4 (109.2) 12.2 (1.7) 20.7 (8.9) 23 Jul (15.8)

CIGP 284.8 (120.7) 19.5 (4.6) 15.0 (6.4) 29 Jul (27.4) 191.3 (131.1) 21.9 (4.6) 8.9 (6.0) 3 Aug (36.3) 277.2 (125.7) 17.5 (3.6) 15.6 (6.0) 24 Jul (26.7)

NEI 360.7 (138.9) 19.9 (3.3) 18.5(7.4) 25 Jun (29.4) 142.9 (106.5) 24.5 (4.1) 6.0 (4.5) 1 Aug (69.6) 352.1 (142.5) 18.6 (3.0) 18.9 (7.1) 25 Jun (29.3)

ENI 221.6 (125.1) 10.9 (1.9) 20.7 (12.0) 21 Jul (25.9) 87.3 (73.5) 14.2 (1.9) 6.3 (5.7) 25 Jul (61.1) 218.4 (127.8) 10.5 (1.4) 20.9 (11.8) 21 Jul (18.8)

INDIA 379.8 (173.2) 19.0 (4.0) 19.7 (8.7) 28 Jul (27.3) 268.4 (185.6) 21.5 (4.1) 12.2 (8.5) 9 Aug (48.3) 356.4 (176.6) 17.7 (3.4) 20.2 (8.4) 23 Jul (24.7)

RAINSW~ , RISW

~ and DURSW~ indicates extreme WS in respect of highest rainfall amount, highest rainfall

intensity and longest duration respectively;

RAINRAINSW

~ , RAINDURSW

~ , RAINRISW

~ and RAINSTARTSW

~ indicates mean rainfall amount, mean duration, mean rainfall intensity and mean starting date respectively of the extreme WS in respect of rainfall amount;

RIRAINSW

~ , RIDURSW

~ , RIRISW

~ and RISTARTSW

~ RAINSTARTSW

~ indicates mean rainfall amount, mean duration, mean rainfall intensity and mean starting date respectively of the extreme WS in respect of rainfall intensity; and

DURRAINSW

~ , DURDURSW

~ , DURRISW

~ and DURSTARTSW

~ indicates mean rainfall amount, mean duration, mean rainfall intensity and mean starting date respectively of the extreme WS in respect of duration.

20

 

TABLE 4. The most extreme wet spell over the 19 subregions during 1951-2007

Sub-regions

In respect of rainfall amount In respect of rainfall intensity In respect of duration

Rainfall RI Dur Period Rainfall RI Dur Period Rainfa

ll RI Dur Period

ESWP 1889.4 30.0 63 25 Jun 61 59.9 59.9 1 19 May 06 1889.4 30.0 63 25 Jun 61

ESEP 1645.2 18.1 91 28 Jun 05 160.4 26.7 6 22 Dec 78 1645.2 18.1 91 28 Jun 05

SCWC 2080.2 49.5 42 24 Jun 61 2037.0 55.1 37 22 Jun 59 1845.9 35.5 52 21 Jun 54

CSEP 1517.9 17.5 90 29 Jun 05 226.7 22.7 10 11 Nov 91 1517.9 17.5 90 29 Jun 05

NCWC 1627.2 37.8 43 29 Jun 54 465.4 46.5 10 16 Jun 53 1313.9 29.9 44 17 Jul 55

SCP 543.3 12.4 44 29 Jun 05 20.6 20.6 1 11 Mar 89 502.2 10.7 47 16 Sep 75

CEC 415.3 20.8 20 5 Oct 58 384.5 38.5 10 6 May 90 296.7 9.3 32 11 Aug 96

NWC 1230.0 31.5 39 11 Jul 83 1042.2 61.3 17 20 Jun 05 993.5 22.1 45 25 Jul 55

NCP 572.3 15.1 38 23 Jul 63 411.8 24.2 17 31 Aug 06 572.3 15.1 38 23 Jul 63

NEC 552.5 15.8 35 17 Jul 92 37.0 37.0 1 12 Nov 02 552.5 15.8 35 17 Jul 92

SNWI 682.5 20.1 34 20 Aug 06 596.5 33.1 18 30 Jul 79 497.2 11.8 42 11 Jul 83

WCI 694.6 16.5 42 2 Jul 56 288.0 24.0 12 21 Aug 87 660.5 14.4 46 29 Jun 59

ECI 1206.7 17.8 68 13 Jun 94 33.7 33.7 1 10 Nov 95 1206.7 17.8 68 13 Jun 94

EIGP 652.5 18.6 35 1 Jul 87 302.9 37.9 8 21 Sep 99 652.5 18.6 35 1 Jul 87

NNWI 490.7 14.0 35 20 Aug 06 264.4 29.4 9 17 May 99 490.7 14.0 35 20 Aug 06

WIGP 563.9 15.2 37 2 Aug 67 147.7 21.1 7 3 Oct 60 547.9 11.9 46 25 Jun 77

CIGP 727.4 19.7 37 1 Jul 55 247.5 35.4 7 9 Aug 87 727.4 19.7 37 1 Jul 55

NEI 904.4 20.1 45 22 May 54 37.8 37.8 1 20 Oct 87 904.4 20.1 45 22 May 54

ENI 640.0 11.0 58 18 Jun 78 22.3 22.3 1 29 Aug 81 567.0 9.3 61 25 Jun 98

21

 

TABLE 5. Mean and standard deviation (SD) of 9 parameters of the dry spells for the 19 sub-regions

Sub regions

DS (SD)

TRAINDS

(SD) in mm TDUR

DS (SD) in days

RIDS

(SD) in mm

RAINDS

(SD) in mmDUR

DS

(SD) in daysSTART

DS

(SD in days) END

DS

(SD in days) age

DS%

(SD) in %

ESWP 5.9(1.9) 509.1 (172.4) 89.5 (31.6) 5.8 (1.0) 86.3 (32.3) 15.2 (6.3) 30 May (21.5) 15 Oct (28.1) 19.4 (7.5)

ESEP 9.7 (2.1) 130.9 (29.5) 172.5 (41.4) 0.8 (0.1) 13.5 (4.7) 17.8 (7.2) 24 Mar (33.2) 6 Dec (17.2) 14.9 (6.7)

SCWC 4.0 (1.4) 459.4 (219.5) 50.7 (31.5) 9.8 (2.1) 114.9 (48.0) 12.7 (6.1) 13 Jun (16.9) 11 Sep (30.5) 15.1 (8.1)

CSEP 8.0 (2.2) 144.8 (32.5) 135.5 (34.6) 1.1 (0.2) 18.1 (5.9) 16.9 (6.9) 8 May (32.3) 26 Nov (17.6) 17.8 (6.0)

NCWC 3.8 (1.2) 358.7 (149.8) 53.3 (29.6) 7.2 (1.5) 94.4 (48.2) 14.0 (8.3) 20 Jun (14.8) 13 Sep (30.0) 16.6 (7.6)

SCP 6.3 (1.9) 119.1 (40.4) 92.1 (31.5) 1.3 (0.3) 18.9 (9.2) 14.6 (7.0) 3 Jun (27.5) 17 Oct (24.6) 17.5 (6.8)

CEC 7.3 (1.9) 221.8 (77.1) 108.1 (44.4) 2.2 (0.5) 30.4 (12.3) 14.8 (7.8) 28 May (36.3) 31 Oct (21.3) 22.3 (8.2)

NWC 3.6 (1.3) 291.0 (104.7) 47.6 (17.3) 6.2 (1.2) 80.8 (38.2) 13.2 (6.5) 26 Jun (12.9) 8 Sep (20.8) 16.7 (6.6)

NCP 5.1 (1.7) 175.6 (60.1) 67.7 (35.3) 2.9 (0.9) 34.4 (15.8) 13.3 (7.1) 17 Jun (30.1) 30 Sep (27.3) 18.1 (7.5)

NEC 7.1 (1.7) 302.7 (103.0) 78.7 (33.7) 4.0 (0.8) 42.6 (17.7) 11.1 (4.7) 14 Jun (25.4) 14 Oct (22.4) 21.6 (7.6)

SNWI 3.1 (1.5) 69.3 (36.6) 53.4 (29.0) 1.5 (0.7) 22.4 (11.7) 17.2 (11.5) 29 Jun (17.8) 10 Sep (32.4) 11.7 (6.3)

WCI 3.8 (1.3) 143.7 (62.9) 44.9 (32.3) 3.8 (1.3) 37.8 (17.7) 11.8 (8.9) 1 Jul (21.9) 13 Sep (32.5) 13.7 (6.3)

ECI 4.6 (1.6) 191.3 (82.6) 47.6 (30.7) 4.4 (1.1) 41.6 (21.5) 10.3 (5.6) 28 Jun (30.6) 17 Sep (22.8) 15.8 (7.2)

EIGP 5.8 (1.6) 291.2 (84.2) 65.6 (21.4) 4.5 (0.8) 50.2 (18.6) 11.3 (4.0) 14 Jun (23.9) 25 Sep (15.8) 21.0 (7.1)

NNWI 3.1 (1.4) 27.5 (18.6) 58.8 (40.8) 0.6 (0.3) 8.9 (8.8) 19.0 (25.0) 27 Jun (32.4) 7 Sep (34.8) 9.5 (7.5)

WIGP 3.7 (1.5) 102.0 (55.9) 44.5 (33.9) 2.6 (0.9) 27.6 (12.9) 12.0 (6.6) 4 Jul (34.7) 12 Sep (21.7) 14.0 (8.2)

CIGP 5.5 (1.5) 185.1 (66.6) 57.8 (31.9) 3.4 (0.7) 33.7 (11.2) 10.5 (4.3) 21 Jun (31.4) 20 Sep (15.4) 16.5 (6.1)

NEI 8.2 (2.1) 553.1 (129.6) 101.0 (30.2) 5.6 (0.8) 67.5 (20.8) 12.3 (4.3) 28 Apr (26.9) 5 Oct (23.3) 26.8 (6.8)

ENI 4.7 (1.6) 173.7 (86.8) 80.9 (61.3) 2.8 (1.1) 37.0 (15.1) 17.2 (11.3) 23 May (61.4) 13 Sep (31.1) 20.3 (10.0)

INDIA 5.4 (1.6) 234.2 (84.9) 76 (33.8) 3.7 (0.9) 45.3 (19.5) 14 (7.9) 8 Jun (28) 2 Oct (25) 17 (7.3)

DS indicates mean number, TRAINDS mean total rainfall and TDUR

DS mean total duration of the DSs;

RIDS mean rainfall intensity, RAIN

DS mean rainfall amount and DURDS mean duration of the individual

DS; STARTDS mean starting date of the first DS; END

DS mean ending date of the last DS; and ageDS

%

mean percentage contribution of the DSs rainfall to the annual total.

22

 

TABLE 6. Mean and standard deviation (SD) of parameters of the extreme dry spell in respect of rainfall, rainfall intensity and duration

Sub region

RAINSD~

RISD~

DURSD~

RAINRAINSD

~

(SD) in mm

RAINRISD

~

(SD) in mm/day

RAINDURSD

~

(SD) in days

RAINSTARTSD

~

(SD) in mm

RIRAINSD

~

(SD) in mm

RIRISD

~

(SD) in mm/day

RIDURSD

~

(SD) in days

RISTARTSD

~

(SD in days)

DURRAINDS

(SD) in mm

DURRISD

~

(SD) in mm/day

DURDURSD

~

(SD) in days

DURSTARTSD

~

(SD in days)

ESWP 24.4 (25.7) 5.7 (3.3) 4.9 (7.2) 14 Jul (45.2) 84.3 (96.5) 3.2 (1.3) 22.3 (20.5) 3 Aug (48.2) 193.0 (101.1) 4.8 (1.7) 40.8 (19.3) 13 Aug (39.2)

ESEP 1.2 (1.5) 0.3 (0.5) 5.8 (9.0) 8 Sep (94.0) 2.9 (4.3) 0.1 (0.1) 13.3 (16.6) 23 Aug (111.2) 44.1 (23.1) 0.8 (0.2) 57.7 (24.2) 15 May (38.6)

SCWC 48.7 (34.8) 10.7 (3.7) 5.3 (6.1) 9 Jul (29.1) 154.0 (128.1) 7.5 (2.8) 23.1 (21.0) 19 Jul (37.0) 219.8 (124.5) 9.0 (3.0) 28.1 (20.8) 30 Jul (30.2)

CSEP 2.1 (1.9) 0.6 (0.5) 4.6 (3.8) 11 Sep (67.0) 6.9 (11.5) 0.3 (0.2) 16.2 (19.1) 21 Sep (78.7) 49.3 (28.2) 1.0 (0.4) 47.5 (24.7) 15 Jun (58.7)

NCWC 39.0 (34.8) 7.3 (3.6) 6.3 (6.2) 16 Jul (33.8) 91.9 (90.8) 5.0 (1.7) 19.2 (20.6) 17 Jul (36.7) 173.3 (92.7) 6.5 (2.0) 29.2 (19.4) 4 Aug (25.0)

SCP 4.4 (5.5) 0.9 (0.7) 5.8 (7.1) 22 Aug(42.8) 9.4 (10.1) 0.5 (0.2) 15.3 (13.9) 14 Aug (48.9) 37.7 (20.9) 1.2 (0.6) 32.1 (13.3) 14 Jul (49.8)

CEC 7.1 (6.9) 1.7 (1.2) 5.9 (6.2) 28 Aug(52.4) 17.6 (20.6) 0.8 (0.5) 19.4 (19.8) 22 Aug (68.3) 62.9 (34.3) 1.8 (0.9) 37.7 (21.8) 21 Jul (68.0)

NWC 39.4 (34.3) 6.6 (2.9) 6.3 (6.5) 20 Jul (28.1) 72.1 (50.2) 4.4 (1.3) 16.1 (11.1) 18 Jul (27.9) 144.8 (70.3) 5.9 (1.8) 24.9 (11.4) 26 Jul (22.9)

NCP 14.6 (15.4) 2.5 (1.3) 6.4 (6.9) 6 Aug (34.4) 26.6 (24.1) 1.7 (0.9) 18.5 (20.3) 9 Aug (46.9) 64.9 (35.1) 2.5 (1.2) 31.1 (24.1) 4 Aug (47.8)

NEC 10.9 (9.7) 3.2 (2.0) 4.5 (6.1) 10 Aug(32.8) 28.3 (28.0) 1.9 (1.0) 18.0 (20.0) 31 Aug (47.4) 88.4 (52.6) 3.5 (1.7) 28.7 (18.7) 3 Aug (51.7)

SNWI 12.1 (11.3) 1.6 (1.0) 10.5 (10.4) 27 Jul (33.0) 28.3 (28.0) 1.9 (1.0) 18.0 (20.0) 31 Aug (47.4) 35.4(21.2) 1.3 (0.8) 31.2 (20.3) 27 Jul (24.8)

WCI 14.6 (9.5) 3.9 (1.8) 4.3 (3.0) 4 Aug (26.1) 40.6 (36.6) 2.5 (1.2) 21.7 (27.5) 2 Aug (35.9) 70.8 (44.8) 3.3 (1.6) 27.8 (26.7) 27 Jul (34.2)

ECI 17.3 (21.7) 3.2 (2.0) 4.4 (4.5) 9 Aug (23.3) 39.6 (45.2) 2.6 (1.5) 17.5 (25.5) 5 Aug (44.8) 78.5 (38.8) 4.1 (1.7) 23.8 (23.3) 2 Aug (45.6)

EIGP 14.2 (11.3) 3.7 (1.9) 4.1 (3.3) 2 Aug (33.4) 32.1 (33.6) 2.7 (1.1) 10.9 (9.4) 12 Aug (41.7) 109.2 (44.2) 4.3 (1.3) 26.1 (11.9) 22 Jul (43.4)

NNWI 5.4 (9.1) 0.4 (0.4) 15.7 (25.5) 27 Jul (27.9) 6.6 (9.1) 0.3 (0.3) 26.4 (32.3) 25 Jul (37.2) 13.1 (12.1) 0.4 (0.3) 35.7 (33.0) 6 Jul (56.5)

WIGP 12.5 (10.9) 2.6 (1.4) 5.5 (4.6) 14 Aug(26.1) 29.6 (31.5) 1.7 (0.9) 20.1 (26.4) 8 Aug (44.8) 48.1 (28.1) 2.3 (1.2) 25.9 (25.2) 28 Jul (43.4)

CIGP 8.8 (6.4) 2.7 (1.8) 3.7 (2.3) 1 Aug (29.0) 17.9 (17.7) 1.8 (1.0) 10.0 (11.4) 29 Jul (39.8) 71.2 (33.7) 3.3 (1.1) 24.2 (20.7) 31 Jul (45.1)

NEI 14.0 (10.1) 4.8 (2.7) 3.7 (2.9) 10 Jul (42.0) 50.6 (48.5) 3.0 (1.4) 17.6 (14.5) 15 Jul (74.5) 149.8 (84.7) 4.9 (2.1) 31.9 (14.5) 5 Jul (69.3)

ENI 10.7 (7.6) 3.1 (1.6) 4.7 (8.9) 31 Jul (40.4) 43.0 (46.6) 1.7 (1.0) 36.6 (43.2) 18 Jun (76.1) 78.2 (47.2) 2.4 (1.4) 50.0(43.6) 13 Jun (75.7)

INDIA 15.9 (14.1) 3.4 (1.8) 5.9 (6.9) 4 Aug (39) 41.2 (40.1) 2.3 (1.0) 19 (20.7) 5 Aug (52.3) 91.2 (49.3) 3.3 (1.3) 33.4 (21.9) 18 Jul (45.8)

RAINSD~ , RISD

~ and DURSD~ indicates extreme DS in respect of lowest rainfall amount, lowest rainfall

intensity and longest duration respectively.

RAINRAINSD

~ , RAINDURSD

~ , RAINRISD

~ and RAINSTARTSD

~ indicates mean rainfall amount, mean duration, mean rainfall intensity and mean starting date respectively of the extreme DS in respect of rainfall amount.

RIRAINSD

~ , RIDURSD

~ , RIRISD

~ and RISTARTSD

~ indicates mean rainfall amount, mean duration, mean rainfall intensity and mean starting date respectively of the extreme DS in respect of rainfall intensity.

DURRAINSD

~ , DURDURSD

~ , DURRISD

~ and DURSTARTSD

~ indicates mean rainfall amount, mean duration, mean rainfall intensity and mean starting date respectively of the extreme DS in respect of duration.

23

 

TABLE 7. The most extreme dry spell over the 19 subregions during 1951-2007.

Sub-regions

In respect of rainfall amount In respect of rainfall intensity In respect of duration

Rainfall RI Dur Period Rainfall RI Dur Period Rainfall RI Dur Period

ESWP 0.6 0.3 2 12 Sep 75 0.6 0.3 2 12 Sep 75 483.9 4.5 107 22 Aug 65

ESEP 0.0 0.0 2 24 Dec 65 0.0 0.0 2 24 Dec 65 80.2 0.6 139 6 Mar 79

SCWC 1.8 0.9 2 8 Jul 05 3.1 0.2 17 25 Sep 05 379.7 3.5 108 24 Aug 65

CSEP 0.0 0.0 2 16 Nov 64 0.0 0.0 2 16 Nov 64 179.6 1.3 138 4 May 68

NCWC 0.5 0.5 1 8 Aug 92 0.5 0.5 1 8 Aug 92 229.4 2.1 107 25 Aug 65

SCP 0.0 0.0 1 13 Sep 60 0.0 0.0 8 15 Oct 84 51.8 0.6 82 12 Mar 89

CEC 0.1 0.1 1 13 May 69 1.2 0.1 19 24 Oct 80 53.0 0.4 121 18 Feb 72

NWC 3.4 3.4 1 20 Jun 70 13.7 2.0 7 16 Aug 71 216.1 3.3 65 11 Aug 51

NCP 0.1 0.1 1 1 Aug 92 0.1 0.1 1 1 Aug 92 85.4 0.6 142 2 Feb 05

NEC 0.2 0.2 1 25 Aug 97 2.5 0.2 12 5 Oct 80 140.3 1.2 117 9 Feb 61

SNWI 0.8 0.8 1 10 Jul 00 2.0 0.1 24 25 May 99 24.5 0.3 93 15 Aug 51

WCI 1.3 0.7 2 11 Jul 86 18.5 0.3 72 23 Sep 62 186.5 1.3 140 12 Mar 06

ECI 0.3 0.3 1 21 Aug 84 0.3 0.3 1 21 Aug 84 214.5 1.5 140 14 Feb 07

EIGP 0.9 0.9 1 26 Sep 71 1.3 0.4 3 12 Jun 70 164.5 2.1 80 1 Apr 98

NNWI 0.0 0.0 1 31 Jul 83 0.0 0.0 1 31 Jul 83 56.3 0.3 168 8 Jul 80

WIGP 1.1 1.1 1 24 Sep 58 1.5 0.2 9 26 Sep 85 155.8 1.2 134 13 Feb 07

CIGP 0.1 0.1 1 27 Jun 75 0.1 0.1 1 27 Jun 75 185.8 1.2 156 31 Jan 59

NEI 0.4 0.4 1 29 Jun 82 8.2 0.2 38 22 Oct 88 304.3 3.8 80 18 Aug 55

ENI 1.0 1.0 1 29 Jul 86 13.2 0.2 69 16 Sep 51 140.9 0.9 156 3 Feb 54

24

 

TABLE 8. Number of subregions that show positive/negative tendency/change (significant at 5% level and above) in the mean of the wet spell parameters during recent 30-year period (1978-2007) compared to earlier period 1951-1977, during recent 20-year 1988-2007 compared to 1951-1987 and during recent 10-year 1998-2007 compared to 1951-1997.

Parameter

Tendency/change in mean of recent 30-year 1978-2007 compared to

1951-1977

Tendency/change in mean of recent 20-year 1988-2007 compared to

1951-1987

Tendency/change in mean of recent 10-year 1998-2007 compared to

1951-1997 Positive Negative Positive Negative Positive Negative

Tendency Change Tendency Change Tendency Change Tendency Change Tendency Change Tendency Change

WS 9 0 10 1 14 1 5 1 13 1 6 0

TRAINWS 9 1 10 1 15 1 4 0 12 0 7 0

TDURWS 4 2 15 0 12 0 7 1 9 0 10 0

RIWS 14 3 5 1 15 0 4 1 13 1 6 0

RAINWS 11 3 8 1 9 2 10 0 10 1 9 0

DURWS 8 1 11 0 7 1 12 0 9 0 10 0

STARTWS 7 1 12 1 3 0 16 2 3 0 16 2

ENDWS 6 1 13 0 5 0 14 0 6 0 13 0 Extreme WS in respect of (high) rainfall amount

RAINRAINSW

~ 9 2 10 2 10 2 9 0 10 1 9 0 RAINRISW

~ 12 1 7 0 11 2 8 0 11 1 8 0 RAINDURSW

~ 7 2 12 1 9 1 10 0 8 1 11 0

RAINSTARTSW

~ 8 2 11 0 9 2 10 0 10 0 9 1

Extreme WS in respect of (high) rainfall intensity RIRAINSW

~ 9 1 10 0 9 2 10 0 8 1 11 0 RIRISW

~ 15 6 4 0 14 3 5 1 13 3 6 0 RIDURSW

~ 8 1 11 0 8 1 11 0 6 0 13 0 RISTARTSW

~ 3 0 16 3 5 0 14 0 7 0 12 0

Extreme WS in respect of (long) duration DURRAINSW

~

9 2 10 2 10 1 9 0 13 1 6 0

DURRISW

~

13 2 6 0 8 1 11 0 11 1 8 0

DURDURSW

~

7 1 12 2 8 1 11 0 10 1 9 0

DURSTARTSW

~

8 1 11 0 10 0 9 0 11 0 8 0

 

 

 

 

 

25

 

TABLE 9. Number of subregions that show positive/negative tendency/change (significant at 5% level and above according to Student’s t-test) in the mean of the dry spell parameters during recent 30-year period (1978-2007) compared to earlier period 1951-1977, during recent 20-year 1988-2007 compared to 1951-1987 and during recent 10-year 1998-2007 compared to 1951-1997.

Parameter

Tendency/change in mean of recent 30-year 1978-2007 compared to

1951-1977

Tendency/change in mean of recent 20-year 1988-2007 compared to

1951-1987

Tendency/change in mean of recent 10-year 1998-2007 compared to

1951-1997 Positive Negative Positive Negative Positive Negative

Tendency Change Tendency Change Tendency Change Tendency Change Tendency Change Tendency Change

DS 9 2 10 0 14 1 5 1 13 1 6 0

TRAINDS 9 1 10 2 14 1 5 0 12 2 7 0

TDURDS 12 1 7 0 14 0 5 0 15 3 4 0

RIDS 6 0 13 0 8 0 11 0 4 0 15 0

RAINDS 8 3 11 1 10 1 9 0 11 2 8 0

DURDS 12 4 7 1 12 1 7 0 13 0 6 0

STARTDS 6 2 13 2 3 0 16 3 4 0 15 2

ENDDS 7 0 12 1 4 0 15 1 6 0 13 0 Extreme DS in respect of (low) rainfall amount

RAINRAINSD~

9 1 10 0 7 1 12 0 11 1 8 0

RAINRISD~ 5 1 14 1 9 0 10 0 7 0 12 0 RAINDURSD~ 9 3 10 0 5 1 14 0 11 1 8 0 RAINSTARTSD~ 10 0 9 1 9 0 10 0 9 0 10 1

Extreme DS in respect of (low) rainfall intensity RIRAINSD~ 10 2 9 0 10 2 9 0 12 2 7 0 RIRISD~ 8 0 11 2 11 0 8 0 6 0 13 0

RIDURSD~ 12 4 7 0 10 2 9 0 10 1 9 0 RISTARTSD~ 5 1 14 0 3 0 16 1 4 0 15 0

Extreme DS in respect of (low) duration DURRAINSD~ 7 3 12 0 8 1 11 0 8 1 11 0

DURRISD~ 7 0 12 3 8 1 11 0 7 0 12 0

DURDURSD~ 10 2 9 0 9 0 10 0 12 1 7 0

DURSTARTSD~ 4 0 15 2 4 0 15 0 3 0 16 2

26

 

 

 

Figure 1. Nineteen sub-regions of India for the purpose of present study. Rainfall threshold for identification of wet and dry spells, climatology of sequence of wet and dry spells, duration of wet and dry spells and normal onset (blue arrow) and withdrawal (red arrow) dates of the summer monsoon are shown on the wet-dry spell gratis of each sub-region.

 

27

Fig. 2 Yearly sequence of wet and dry spells over extreme southwest peninsula (ESWP; SR1) during 1951-2007

28

Fig. 3 Yearly sequence of wet and dry spells over extreme southeast peninsula (ESEP; SR2) during 1951-2007

29

Fig. 4 Yearly sequence of wet and dry spells over southern central West Coast (SCWC; SR3) during 1951-2007

30

Fig. 5 Yearly sequence of wet and dry spells over central southeast peninsula (CSEP; SR4) during 1951-2007

31

Fig. 6 Yearly sequence of wet and dry spells over northern central West Coast (NCWC; SR5) during 1951-2007

32

Fig. 7 Yearly sequence of wet and dry spells over south central peninsula (SCP; SR6) during 1951-2007

33

Fig.8 Yearly sequence of wet and dry spells over central East Coast (CEC; SR7) during 1951-2007

34

Fig. 9 Yearly sequence of wet and dry spells over northern West Coast (NWC; SR8) during 1951-2007

35

Fig.10 Yearly sequence of wet and dry spells over northern central peninsula (NCP; SR9) during 1951-2007

36

Fig. 11 Yearly sequence of wet and dry spells over northern East Coast (NEC; SR10) during 1951-2007

37

Fig. 12 Yearly sequence of wet and dry spells over south northwest India (SNWI; SR11) during 1951-2007

38

Fig. 13 Yearly sequence of wet and dry spells over western central India (WCI; SR12) during 1951-2007

39

Fig. 14 Yearly sequence of wet and dry spells over eastern central India (ECI; SR13) during 1951-2007

40

Fig. 15 Yearly sequence of wet and dry spells over eastern Indo-Gangetic plains (EIGP; SR14) during 1951-2007.

41

Fig. 16 Yearly sequence of wet and dry spells over northern northwest India (NNWI; SR15) during 1951-2007

42

Fig. 17 Yearly sequence of wet and dry spells over western Indo-Gangetic plains (WIGP; SR16) during 1951-2007

43

Fig. 18 Yearly sequence of wet and dry spells over central Indo-Gangetic plains (CIGP; SR17) during 1951-2007

44

Fig. 19 Yearly sequence of wet and dry spells over northeast India (NEI; SR18) during 1951-2007

45

Fig. 20 Yearly sequence of wet and dry spells over extreme northern India (ENI; SR19) during 1951-2007

46 

 

0

2

4

6

8

10

121950 1960 1970 1980 1990 2000

SR-1

0

4

8

12

16

1950 1960 1970 1980 1990 2000

SR-2

0

2

4

6

8 SR-3

0

4

8

12

16

20

SR-4

0

2

4

6

8

SR-5

0

4

8

12

SR-6

0

4

8

12

16 SR-7

0

2

4

6

8 SR-8

0

2

4

6

8

10SR-9

0

4

8

12

SR-10

0

2

4

6

8 SR-11

0

2

4

6

8 SR-12

0

2

4

6

8

10

SR-13

0

4

8

12

SR-14

0

2

4

6

8

SR-15

0

2

4

6

8

10

SR-16

0

2

4

6

8

10

SR-17

1950 1960 1970 1980 1990 20000

4

8

12

16 SR-18

1950 1960 1970 1980 1990 20000

2

4

6

8

10

SR-19

.

Year

Year

Num

ber

of w

et s

pel

ls

Figure 21. Interannual variation (histogram) of number of wet spells over the 19 sub-regions. The continuous curve shows 5-point binomial low-pass filtered values.

47 

 

22

24

26

28

30

32

34

1951 1961 1971 1981 1991 2001

SR-1

4

8

12

1951 1961 1971 1981 1991 2001

SR-2

30

40

SR-3

8

12

16

SR-4

20

25

30

35

40

SR-5

9

12

SR-6

8

12

16 SR-7 40 SR-8

8

12

16

SR-9

15

18SR-10

Year

Year

Rai

nfa

ll in

tensi

ty p

er w

et d

ay (

mm

/day

)

10

20SR-11

12

15

18

SR-12

15

18SR-13

15

18

21

SR-14

10

20

SR-15

9

12

15

SR-16

12

16

20

24

SR-17

1951 1961 1971 1981 1991 2001

15

20

SR-18

1951 1961 1971 1981 1991 2001

10

12

14

SR-19

.

Figure 22. Interannual variation (thin curve) and 5-point binomial low-pass filtered values (thick curve) of rainfall intensity during wet spells over the 19 sub-regions.

48 

 

10

20

301951 1961 1971 1981 1991 2001

SR-1

4

8

12

16

20

1951 1961 1971 1981 1991 2001

SR-2

10

15

20

25 SR-3

8

12

16

20

SR-4

5

10

15

20

25

SR-5

5

10

15 SR-6

6

9

12 SR-7

5

10

15

20

25

SR-8

10

20

30

SR-9

6

9

12

15

SR-10

Year

Year

Mea

n D

ura

tion o

f in

div

idual

wet

spel

l (d

ays)

5

10

15

20SR-11

10

20

30

SR-12

7

14

21 SR-13

6

9

12 SR-14

10

20

30

SR-15

5

10

15

20

SR-16

4

8

12 SR-17

1951 1961 1971 1981 1991 2001

6

9

12

15

SR-18

1951 1961 1971 1981 1991 2001

4

6

8

10

12

14

SR-19

.

 

Figure 23. Interannual variation (thin curve) and 5-point binomial low-pass filtered values (thick curve) of average duration of individual wet spell over the 19 sub-regions.

49 

 

100

120

140

160

1801951 1961 1971 1981 1991 2001

SR-1

40

80

120

1601951 1961 1971 1981 1991 2001

SR-2

125

150

SR-3

50

100

150

200

SR-4

140

160

SR-5

100

150

200 SR-6

50

100

150

SR-7

160

180 SR-8

50

100

150

200 SR-9

50

100

150

200SR-10

Year

Year

Sta

rtin

g d

ate

of

the

firs

t w

et s

pel

l (d

ay-i

n-y

ear)

160

200

SR-11

100

150

200 SR-12

50

100

150

200 SR-13

100

150

200 SR-14

100

150

200 SR-15

50

100

150

200 SR-16

50

100

150

200

SR-17

1951 1961 1971 1981 1991 2001

50

100

150 SR-18

1951 1961 1971 1981 1991 2001

50

100

150

200

SR-19

.

Figure 24. Interannual variation (thin curve) and 5-point binomial low-pass filtered values (thick curve) of starting date of first wet spell over the 19 sub-regions.

50 

 

500

1000

1500

20001951 1961 1971 1981 1991 2001

SR-1

0

500

1000

1500

1951 1961 1971 1981 1991 2001

SR-2

500

1000

1500

2000 SR-3

500

1000

1500

SR-4

400

800

1200

SR-5

150

300

450

600

SR-6

150

225

300

375SR-7

300

600

900

1200 SR-8

125

250

375

500 SR-9

150

300

450

SR-10

Year

Year

Rai

nfa

ll am

ount

of

an e

xtre

me

wet

spel

l (i

n r

espec

t of

rain

fall)

(m

m)

150

300

450

SR-11

200

400

600

SR-12

300

600

900

1200

SR-13

150

300

450

600SR-14

0

150

300

450SR-15

200

400

600

SR-16

200

400

600

800SR-17

1951 1961 1971 1981 1991 2001

200

400

600

800SR-18

1951 1961 1971 1981 1991 20010

200

400

600 SR-19

.

Figure 25. Interannual variation (thin curve) and 5-point binomial low-pass filtered values (thick curve) of rainfall amount of extreme wet spell in respect of rainfall amount over the 19 sub-regions.

51 

 

30

40

50

60

1951 1961 1971 1981 1991 2001

SR-1

10

20

301951 1961 1971 1981 1991 2001

SR-2

30

40

50 SR-3

10

20 SR-4

20

30

40 SR-5

10

15

20

SR-6

10

20

30

40

SR-7

30

40

50 SR-8

8

12

16

20

24

SR-9

20

30SR-10

Year

Year

Rai

nfa

ll in

tensi

ty o

f an

ext

rem

e w

et s

pel

l (i

n r

espec

t of

RI)

(m

m/d

ay)

10

20

30 SR-11

15

20

25

SR-12

20

30 SR-13

20

30SR-14

0

10

20

30

SR-15

10

15

20 SR-16

20

30

40 SR-17

1951 1961 1971 1981 1991 2001

20

30SR-18

1951 1961 1971 1981 1991 2001

10

15

20 SR-19

.

Figure 26. Interannual variation (thin curve) and 5-point binomial low-pass filtered values (thick curve) of rainfall intensity of extreme wet spell in respect of rainfall intensity over the 19 sub-regions.

52 

 

15

30

45

60

1951 1961 1971 1981 1991 2001

SR-1

25

50

75

1001951 1961 1971 1981 1991 2001

SR-2

15

30

45SR-3

25

50

75 SR-4

15

30

45 SR-5

0

15

30

45 SR-6

10

20

30 SR-7

15

30

45 SR-8

15

30

SR-9

10

20

30 SR-10

Year

Year

Mea

n D

ura

tion o

f an

ext

rem

e w

et s

pel

l (i

n r

espec

t of

dura

tion)

(day

s)

15

30

SR-11

10

20

30

40

SR-12

20

40

60 SR-13

10

20

30SR-14

10

20

30 SR-15

15

30

45 SR-16

10

20

30

40

SR-17

1951 1961 1971 1981 1991 2001

15

30

45 SR-18

1951 1961 1971 1981 1991 20010

20

40SR-19

.

Figure 27. Interannual variation (thin curve) and 5-point binomial low-pass filtered values (thick curve) duration of extreme wet spell in respect of duration over the 19 sub-regions.

53 

 

4

6

8

1951 1961 1971 1981 1991 2001

SR-1

0.6

0.8

1.0

1.21951 1961 1971 1981 1991 2001

SR-2

10

SR-3

0.5

1.0

1.5 SR-4

4

6

8

10

11

SR-5

1.0

1.5

2.0

SR-6

1

2

3 SR-7

3

6

9

SR-8

1

2

3

4

5

SR-9

2

3

4

5

6

SR-10

Year

Year

Rai

nfa

ll in

tensi

ty p

er d

ry d

ay (

mm

/day

)

0

1

2

SR-11

2

4

6

SR-12

0

2

4

6 SR-13

2

3

4

5

6

SR-14

0

1 SR-15

2

4 SR-16

1

2

3

4

5

SR-17

1951 1961 1971 1981 1991 20014

6

8

SR-18

1951 1961 1971 1981 1991 20010

1

2

3

4 SR-19

.

Figure 28. Interannual variation (thin curve) and 5-point binomial low-pass filtered values (thick curve) of rainfall intensity during dry spells over the 19 sub-regions.

54 

 

10

20

30

1951 1961 1971 1981 1991 2001

SR-1

10

20

30

40

50

1951 1961 1971 1981 1991 2001

SR-2

10

20

30 SR-3

10

20

30

40

SR-4

10

20

30

40

SR-5

10

20

30

40

SR-6

10

20

30

40 SR-7

10

20

30

40

SR-8

10

20

30

40

SR-9

10

20SR-10

Year

Year

Mea

n D

ura

tion o

f in

div

idual

dry

spel

l (d

ays)

0

10

20

30

40

50 SR-11

10

20

30

40

50

SR-12

7

14

21

28

SR-13

10

20 SR-14

50

100

150 SR-15

10

20

30

40

SR-16

10

20

30 SR-17

1951 1961 1971 1981 1991 2001

10

20

30

SR-18

1951 1961 1971 1981 1991 2001

10

20

30

40

50

SR-19

.

Figure 29. Interannual variation (thin curve) and 5-point binomial low-pass filtered values (thick curve) of duration of individual dry spell over the 19 sub-regions.

55 

 

0

50

100

1951 1961 1971 1981 1991 2001

SR-1

0

5

1951 1961 1971 1981 1991 2001

SR-2

50

100

150 SR-3

0

5

SR-4

0

50

100

150

200

250

SR-5

0

10

20

30

40

SR-6

0

20

40

SR-7

0

75

150SR-8

0

30

60SR-9

0

20

40

60 SR-10

Year

Year

Rain

fall

am

ount

of

an e

xtre

me

dry

spel

l (i

n r

espec

t of

Rai

nfa

ll) (

mm

)

0

20

40 SR-11

0

20

40

60

SR-12

0

50

100 SR-13

0

20

40

60

SR-14

0

20

40

60

SR-15

0

20

40

60

SR-16

0

10

20

30

SR-17

1951 1961 1971 1981 1991 20010

20

40 SR-18

1951 1961 1971 1981 1991 20010

10

20

30 SR-19

.

Figure 30. Interannual variation (thin curve) and 5-point binomial low-pass filtered values (thick curve) of rainfall amount of extreme dry spell in respect of rainfall amount over the 19 sub-regions.

56 

 

0

2

4

6

1951 1961 1971 1981 1991 2001

SR-1

0.0

0.2

0.4

0.6

1951 1961 1971 1981 1991 2001

SR-2

5

10

SR-3

0.0

0.5

1.0SR-4

0

2

4

6

8 SR-5

0.5

1.0

SR-6

0

1

2 SR-7

3.0

4.5

6.0SR-8

0

1

2

3

4

5

SR-9

0

1

2

3

4

5

SR-10

Year

Year

Rai

nfa

ll in

tensi

ty o

f ex

trem

e dry

spel

l (i

n r

espec

t of

RI)

(m

m/d

ay)

0

1

2

SR-11

0

2

4

6

SR-12

0

2

4

6 SR-13

0

1

2

3

4

5

6 SR-14

0.0

0.5

1.0 SR-15

0

1

2

3

4

SR-16

0

1

2

3

4

5

SR-17

1951 1961 1971 1981 1991 20010

2

4

SR-18

1951 1961 1971 1981 1991 20010

1

2

3

4 SR-19

.

Figure 31. Interannual variation (thin curve) and 5-point binomial low-pass filtered values (thick curve) of rainfall intensity of extreme dry spell in respect of rainfall intensity over the 19 sub-regions.

57 

 

50

100

1951 1961 1971 1981 1991 2001

SR-1

50

100

1501951 1961 1971 1981 1991 2001

SR-2

50

100 SR-3

50

100

150

SR-4

50

100SR-5

20

40

60

80

SR-6

50

100SR-7

20

40

60 SR-8

50

100

150

SR-9

50

100SR-10

Year

Year

Mea

n D

ura

tion o

f an

ext

rem

e dry

spel

l (i

n r

espec

t of

dura

tion)

(day

s)

40

80 SR-11

50

100

SR-12

50

100SR-13

25

50

75

SR-14

0

50

100

150 SR-15

50

100SR-16

0

50

100

150

SR-17

1951 1961 1971 1981 1991 2001

30

60SR-18

1951 1961 1971 1981 1991 2001

50

100

150

200

SR-19

.

Figure 32. Interannual variation (thin curve) and 5-point binomial low-pass filtered values (thick curve) of duration of extreme dry spell in respect of duration over the 19 sub-regions.

58 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

FIG. 33. Map showing tendency/change (at 5% level and above) in mean of the indicated parameter of the wet spell and its extremes during the 20-year period 1988-2007 compared to the 1951-1987. Grey shading indicates positive tendency/change and black negative tendency/change. Letter ‘S’ denotes that the change is significant at 5% level and above (Student’s t-test).               

59 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

FIG. 34. Same as in Figure 14, but for the dry spells.